Development of a non-invasive nanoparticle-based vaccination strategy protecting against influenza infection

Von der Fakultät für Lebenswissenschaften

der Technischen Universität Carolo-Wilhelmina zu Braunschweig

zur Erlangung des Grades eines

Doktors der Naturwissenschaften

(Dr. rer. nat.)

genehmigte

D i s s e r t a t i o n

von Simon Marc Pascal Delandreaus Lisieux, Frankreich

1. Referent Prof. Dr. Stefan Dübel 2. Referent Prof. Dr. Carlos A. Guzmán

eingereicht am: 11.09.2019

mündliche Prüfung (Disputation) am: 09.12.2019

Druckjahr 2020

Vorveröffentlichungen der Dissertation

Teilergebnisse aus dieser Arbeit wurden mit Genehmigung der Fakultät für Lebenswissenschaften, vertreten durch den Mentor der Arbeit, in folgenden Beiträgen vorab veröffentlicht:

Posterbeitrag

Delandre S, Kliesch L, Nasr S, Loretz B, Schulze K, Müller A, Guzmán CA, Lehr CM. Nanocarriers for minimally invasive intradermal vaccination (Poster 46). HIPS Symposium, Saarbrücken (2019).

To my parents

i

Acknowledgements

Throughout the past three and a half years working on this thesis, I have received many supports that I would like to mention here. First and foremost, I would like to thank Prof. Dr. Carlos A. Guzman for giving me the opportunity to join his research group and to work on this interesting and challenging project. Thanks to him, I could expand my technical knowledge and improve my scientific skills.

I also would like to express my gratitude to Prof. Dr. Claus-Michael Lehr and Prof. Dr. Andreas Müller for being in my thesis committee and for their great scientific input. I would also like to thanks Prof. Dr. Stefan Dübel for accepting to be my mentor.

This work would not have been possible without the help and the supervision of Kai Schulze, with whom I got to know the field of skin vaccination and worked side by side during those three years. I should also mention the other members of the VAC group: Thomas Ebensen, Peggy Riese, Steffi Trittel and Sarah Hennig for the scientific part, but also Elena Reinhard, Ulrike Bröder, and Hanna Shkarlet for the technical aspect. Thank you also to Cornelia Senske and to Blair Prochnow.

Living in Braunschweig also gave me the opportunity to meet amazing people with whom I spent awesome moments, among them I should mention Hanna (again), Shiwani who introduce me to the joy of doing a Ph.D., Charlotte, Nathiana, Christian, Kevin and of course Sven. Even with new friends, I do not forget my beloved friends from France, Stephanie, Yoann, Delphine and the famous Secotine that followed me since my very first steps in science.

And of course, last but definitely not least, I would like to acknowledge my parents, Jean-Claude and Corinne, as well as my sisters, Virginie and Emilie, and my nephews Clara and Louis for their constant support despite the distance.

ii

Table of contents

Acknowledgements .............................................................................................................. i

Table of contents ................................................................................................................. ii

Table of figures .................................................................................................................. vii

Table of tables ..................................................................................................................... x

Abbreviations ...................................................................................................................... xi

Summary .............................................................................................................................. 1

Zusammenfassung ............................................................................................................... 2

Introduction ......................................................................................................................... 3

1. Vaccination ............................................................................................................... 3

a. The ins and outs of vaccination .............................................................................. 3

b. Challenges for vaccination ..................................................................................... 4

2. Skin vaccination ....................................................................................................... 5

a. Skin anatomy .......................................................................................................... 5

b. Skin Immune System (SIS) and Skin-Associated Lymphoid Tissue (SALT) ....... 6

c. Transfollicular vaccination ................................................................................... 10

d. Targeting of the skin dendritic cells ..................................................................... 13

3. Vaccine formulation for transdermal application ................................................... 13

a. Use of inactivated viruses as vaccine antigens ..................................................... 13

b. Nanoparticles as non-viral vehicles ...................................................................... 14

iii

c. Adjuvants .............................................................................................................. 15

d. Influenza virus ...................................................................................................... 16

Aim of the thesis ................................................................................................................ 19

Materials ............................................................................................................................ 20

1. Instruments ............................................................................................................. 20

2. Chemical compounds and reagents ........................................................................ 21

3. Antibodies ............................................................................................................... 23

4. Solutions, buffers and culture media ...................................................................... 25

5. Kit for cytokine detection and cell enrichment ...................................................... 26

6. Antigens, adjuvants and nanocarriers ..................................................................... 27

7. Influenza virus ........................................................................................................ 27

Methods ............................................................................................................................. 29

1. Immunization protocol ........................................................................................... 29

2. Sample collection and processing .......................................................................... 29

a. Spleen ................................................................................................................... 29

b. Skin-draining lymph nodes .................................................................................. 30

c. Skin tissue ............................................................................................................. 30

3. Determination of cell concentrations ..................................................................... 30

4. Detection of cytokine-producing cells using the ELISpot assay ............................ 31

5. Antigen-specific multifunctional T-cells analysis .................................................. 31

6. Cell proliferation assay ........................................................................................... 31

7. Multiplex FlowCytomix (Cytometric Bead Array (CBA)) .................................... 32

iv

8. Cell staining for flow cytometric analysis .............................................................. 32

9. Antigen-specific antibody detection and functionality assessment ........................ 35

a. IgG and IgG subclasses ELISA ............................................................................ 35

b. Hemagglutination Inhibition assay (HAI) ............................................................ 35

c. Enzyme-Linked Lectin Assay (ELLA) ................................................................ 36

10. Naïve T-lymphocytes magnetic enrichment ........................................................... 36

11. Adoptive transfer .................................................................................................... 37

12. In vivo determination of cytotoxic T-lymphocytes (CTL) ..................................... 37

13. Challenge of vaccinated mice ................................................................................. 38

14. In vivo Imaging System (IVIS) .............................................................................. 38

15. Skin explant culture ................................................................................................ 38

16. In vivo imaging using 2-Photon Microscopy ......................................................... 39

17. Statistical analysis .................................................................................................. 39

Results ............................................................................................................................... 40

1. Vaccination strategy ............................................................................................... 40

a. Skin integrity before immunization ...................................................................... 40

b. No oral “cross-vaccination” by grooming ............................................................ 41

2. Proof of concept: Efficient non-invasive transfollicular delivery of inactivated influenza virus as a vaccine antigen .............................................................................. 42

a. Adoptive transfer model ....................................................................................... 42

b. Characterization of the adaptive immune responses stimulated following vaccination with non-adjuvanted inactivated H1N1. ................................................. 43

v

c. Tailored immune response using adjuvanted formulations.................................. 47

d. Adjuvanted H5N1 formulation improve strength and quality of the immune response even when using low antigen doses ............................................................ 51

3. Can polymeric nanoparticles be utilized for the transfollicular vaccination route: the polyphosphazene polymer ............................................................................................. 57

a. The nanocarrier penetrate into the hair follicle .................................................... 57

b. Successful delivery of the vaccine antigen to skin dendritic cells ....................... 58

c. Transfollicular immunization using polyphosphazenes as nanocarrier stimulates antigen-specific immune responses ........................................................................... 59

d. Transfollicular immunization using formulations based on polyphosphazenes protects mice against live virus challenge .................................................................. 64

4. Polyphosphazene polymer based formulation for the recombinant HA protein .... 65

a. Recombinant HA protein efficiently activate antigen-specific CD4+ and CD8+ T-cells when applied via the TF route ........................................................................... 65

b. Characterization of the adaptive immune responses ............................................ 67

c. Transfollicular immunization using polyphosphazene polymer formulated with recombinant HA protein protects mice against influenza infection ........................... 72

5. Deciphering the local early events triggered by transfollicular immunization. ..... 74

a. Antigen uptake by skin DC following TF immunization ..................................... 74

b. DC activation following transfollicular immunization ........................................ 77

c. Migration of skin dendritic cells. ......................................................................... 78

d. Migration of skin dendritic cells to the draining lymph nodes ............................ 81

e. Deeper characterization of the migrating skin dendritic cells .............................. 82

f. Local pro-inflammatory chemokine profile ......................................................... 83

vi

g. Neutrophils recruitment in the skin ...................................................................... 84

Discussion .......................................................................................................................... 86

Conclusion and Outlook .................................................................................................... 96

References ......................................................................................................................... 98

vii

Table of figures

Figure 1: Skin anatomy. ..................................................................................................... 5

Figure 2: Mouse skin dendritic cell subsets. ...................................................................... 6

Figure 3: Human and mouse skin immune cells. ............................................................... 8

Figure 4: Mouse dendritic cells and their T-cells polarization features. .......................... 10

Figure 5: Vaccine delivery techniques for cutaneous vaccination. .................................. 11

Figure 6: Gating strategy for skin dendritic cell subsets. ................................................. 33

Figure 7: Gating strategy for skin migratory dendritic cells in the lymph nodes ............ 34

Figure 8: Skin integrity post depilation indicated by neutrophil influx into the skin following depilation of murine skin. ................................................................................. 40

Figure 9: No oral cross-vaccination by grooming following transfollicular vaccination. 41

Figure 10: HA-specific proliferation of CD4+ and CD8+ lymphocytes stimulated by c-di-AMP adjuvanted inactivated H1N1 virus. ......................................................................... 43

Figure 11: Systemic humoral immune responses induced in mice after three immunizations with inactivated H1N1 virus via TF route. ............................................... 44

Figure 12: Cellular immune response stimulated by non-adjuvanted inactivated H1N1 virus following TF vaccination. ........................................................................................ 45

Figure 13: Cytokine profile (IL-2, IL-4, IL-17) observed in mice after three immunizations with non-adjuvanted H1N1 virus antigen via TF route. .................................................... 46

Figure 14: Protection of vaccinated animals against influenza A virus infection. ........... 47

Figure 15: Systemic humoral immune responses induced in mice after three immunizations with adjuvanted H1N1 virus antigen via TF route. .................................. 48

Figure 16: Anti-HA and anti-NA neutralizing antibody titers in sera.............................. 49

viii

Figure 17: Cellular immune responses stimulated by adjuvanted inactivated H1N1 virus. ........................................................................................................................................... 50

Figure 18: Cytokine profile (IL-2, IL-4, IL-17) observed in mice after three immunizations with adjuvanted H1N1 virus antigen via TF route. ........................................................... 51

Figure 19: Systemic humoral immune responses induced in mice after three immunizations with adjuvanted H5N1 virus antigen via TF route. .................................. 53

Figure 20: Cellular immune response stimulated by the adjuvanted inactivated H5N1 virus. .................................................................................................................................. 54

Figure 21: Cytokine profile (IFNγ, IL-2, IL-4 and IL17) observed in mice after three immunizations with adjuvanted H5N1 virus antigen via TF route. .................................. 56

Figure 22: In vivo 2-Photon microscopy imaging. ........................................................... 58

Figure 23: OVA-specific proliferation of CD4+ and CD8+ lymphocytes. ....................... 59

Figure 24: Systemic humoral immune responses induced in mice after four immunizations with polyphophazene polymers encompassing OVA protein via TF route. ..................... 60

Figure 25: Cellular immune response stimulated by c-di-AMP adjuvanted OVA-polyphosphazene formulations. ......................................................................................... 61

Figure 26: Cytokine profile (IFNγ, IL-2, IL-4, and IL17) observed in mice after four immunizations with polyphophazene polymers encompassing OVA protein via TF route ........................................................................................................................................... 63

Figure 27: OVA-specific cytotoxic lymphocytes killing assay. ...................................... 64

Figure 28: Protection of vaccinated animals against influenza A virus infection ............ 65

Figure 29: HA-specific proliferation of CD4+ and CD8+ lymphocytes following TF vaccination using recombinant HA protein co-formulated with either of the two polyphosphazene polymers (PCPP or PCEP).................................................................... 67

Figure 30: Systemic humoral immune responses induced in mice after four immunizations with polyphosphazene polymers encompassing recombinant HA protein via TF route ... 68

ix

Figure 31: Cellular immune response stimulated by adjuvanted HA-polyphosphazene formulations. ...................................................................................................................... 69

Figure 32: Antigen-specific cytokine-profiles stimulated by HA-polymers co-administered with different adjuvants. .............................................................................. 71

Figure 33: Protection of vaccinated animals against influenza A virus infection. ........... 73

Figure 34: In vivo bioluminescence imaging of mice at day 7 post-inoculation of live H1N1 virus expressing the luciferase protein ................................................................... 74

Figure 35: Immunization and sampling schedule............................................................. 75

Figure 36: Kinetic of vaccine antigen uptake and processing by skin DC ...................... 76

Figure 37: Vaccine antigen uptake and processing by skin DC subpopulations ............. 77

Figure 38: Activation of skin DC after TF immunization of mice with different PCPP based formulations. ............................................................................................................ 78

Figure 39: Migration kinetics of the four main skin DC subpopulations......................... 79

Figure 40: Impact of the formulation on the frequency of the skin DC populations. ...... 80

Figure 41: Recruitment of migratory skin DC following TF immunization. ................... 81

Figure 42: Migration of antigen-positive DC subpopulations. ........................................ 83

Figure 43: Pro-inflammatory chemokine profiles in the skin following TF vaccination monitored for 96 h.. ........................................................................................................... 84

Figure 44: Kinetic of neutrophils recruitment in the skin following TF immunization ... 85

x

Table of tables

Table 1: Clinical and pre-clinical skin vaccination………………………………………12

Table 2: Adjuvants licensed for human use……………………………………………..16

Table 3: Instruments and consumables……………………….…………………………20

Table 4: Chemical compounds and reagents……………….…………….……………...21

Table 5: Antibodies used for flow cytometry…………….…………….………………..23

Table 6: Antibodies used for Enzyme-Linked Immunospot Assay (ELISpot)….………24

Table 7: Antibodies used for Enzyme-Linked Immunosorben Assay (ELISA)…….…..25

Table 8: Solutions, buffers and culture media…………………………….…………….25

Table 9: Kits for cytokine detection and cell enrichment…………..……….…………..26

Table 10: Antigens, adjuvants and nanocarriers……………………………….………..27

Table 11: Influenza strains used for animal challenge………………………….………27

xi

Abbreviations

Abbreviation Full name

ABTS 2,2´-azino-bis (3-ethylbenzthiazoline-6 sulfonic acid) diammonium salt

ACK Ammonium chloride-potassium

AEC 3-amino-9-ethyl-carbazole (AEC substrate kit)

α-GalCerMPEG = αGCMPEG

a-galactosylceramide methyl polyethylene glycol

APCs Antigen-presenting cells

AUC Area under the curve

BSA Bovine serum albumin

CD Cluster of differentiation

c-di-AMP Bis-(3´, 5´)-cyclic dimeric adenosine monophosphate

CFSE Carboxyfluorescein diacetate succinimidyl ester

Ci Curie

CLR C-type lectin receptor

ConA Concanavalin A

Cpm Counts per minute

CSSS Cyanoacrylate Skin Surface Stripping

xii

CTL Cytotoxic T-lymphocyte

CXCL Chemokine ligand

CXCR Chemokine receptor

DCs Dendritic cells

DMSO Dimethyl sulfoxide

DNA Deoxyribonucleic acid

DN cDC Double Negative conventional Dendritic cells

EDTA Ethylenediaminetetraacetic acid

e.g. Exempli gratia

ELISA Enzyme-linked immunosorbent assay

ELISPOT Enzyme-linked immunospot assay

ELLA Enzyme-Linked Lectin Assay

Eq/ml Equivalent from total inactivated virus per milliliter

FACS Fluorescence activated cell sorting

FCA Freund’s complete adjuvant

FCS Fetal calf serum

FDA Food and Drug Administration

FITC Fluorescein-isothiocyanate

xiii

GM-CSF Granulocyte-macrophage colony stimulating factor

HA Hemaglutinin

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HF Hair follicle

h.i. Heat inactivated

HRP Horseradish peroxidase

i.d. Intradermal

i.e. Id est

IFN Interferon

Ig Immunoglobulin

IL Interleukin

i.m. Intramuscular

i.v. Intravenously

IVIS In vivo Imaging System

LC Langerhans cells

LNs Lymph nodes

MALT Mucosa-Associated Lymphoid Tissue

MFI Mean Fluorescence Intensity

xiv

MHC Major histocompatibility complex

min Minute

NA Neuraminidase

NK Natural killer cells

NKT Natural killer T-cells

NPs Nanoparticles

OVA Ovalbumin

PAMP Pathogen-associated molecular pattern

PBS Phosphate-buffered saline

PE Phycoerythrin

PerCP Peridinin-chlorophyll-protein complex

PE-Cy7 Phycoerythrin-cyanin 7

PFA Paraformaldehyde

PLGA Poly lactic-co-glycolic acid

PMA Phorbol 12-myrisate 13-acetate

RNA Ribonucleic acid

RORγt RAR-related orphan receptor gamma

RT Room temperature

xv

SALT Skin-Associated Lymphoid Tissue

SC Stratum Corneum

s.c. Subcutaneous

SEM Standard error of the mean

STING Stimulator of interferon genes

TCR T-cell receptor

TCZ T-cell zone

TF Transfollicular

Tfh T-follicular helper cells

Th T-helper

Th1 T-helper 1

Th2 T-helper 2

Th17 T-helper 17

TLR Toll-like receptor

TSLP Thymic Stromal Lymphopoietin

UV Ultraviolet

XCR1 X-C Motif Chemokine Receptor 1

WT Wild type

1

Summary

Nowadays, vaccination represents the most effective medical intervention to control and prevent the spreading of infectious diseases in the population. Nonetheless, the emergence of new pathogens, the logistic constraints for the vaccine storage and the safety concerns regarding the use of whole pathogens, as well as the rising wave of skepticism among the population toward vaccination, have led to the rebound of some almost eradicated pathogens and delayed the success to control some other diseases like influenza infection. Hence, the development of non-invasive needle-free vaccination strategies combined with subunit vaccines able to stimulate protective immune responses is required. In this thesis, we have confirmed the potential of the transfollicular (TF) route for the delivery and the generation of a protective immune response using inactivated influenza viruses. Moreover, by investigating the early innate immune events occurring in the skin after TF immunization using polymeric nanocarriers, we have shown that the nanocarrier is needed for the effective delivery and uptake of the vaccine antigen. In addition, the co-formulation of the nanoparticles with either of Bis-(3´, 5´)-cyclic dimeric adenosine monophosphate (c-di-AMP) or pegylated αGalactosylceramide (αGCMPEG) adjuvants, help to tailor the innate immune response by acting on specific skin dendritic cells and therefore participate in the induction of a differentially polarized adaptive immune response. Altogether, this study has revealed not only the potential of the TF route for vaccine delivery but also that the combination of this route with adjuvanted polymeric nanoparticles is able to induce a protective immune response against live influenza virus challenge.

2

Zusammenfassung

Impfungen stellen die effizienteste Erfindung in der Medizin dar, um Infektionskrankheiten zu kontrollieren und ihre Ausbreitung zu verhindern bzw. einzudämmen. Nichtsdestotrotz führen neu auftretende Krankheiten, Sicherheitsbedenken bei der Verwendung von Lebendimpfstoffen, sowie logistische Probleme bei der Lagerung und Einhaltung von Kühlketten und nicht zuletzt die zunehmende Impfmüdigkeit dazu, dass bereits unter Kontrolle oder gar ausgerottet geglaubte Pathogene wieder neu erstarken. Aus diesen Gründen ist es notwendig, neue nadelfreie nicht-invasive Impfstrategien zu entwickeln, die sich durch ein erhöhtes Sicherheitsprofil bei gleichzeitig hoher Effektivität kennzeichnen. In der vorliegenden Arbeit konnten wir bestätigen, dass sich die transfollikuläre (TF) Impfroute dazu eignet, schützende Immunantworten gegen das Influenzavirus zu stimulieren. Darüber hinaus konnte durch Analyse der unmittelbar nach TF Immunisierung ablaufenden Prozesse des angeborenen Immunsystems in der Haut gezeigt werden, dass Nanopartikel für diese Impfstrategie notwendig sind, um einen effektiven Transport des Impfstoffantigens zu den, sowie dessen Verarbeitung durch die Immunzellen zu gewährleisten. Ferner kann durch Zugabe der Adjuvantskandidaten Bis-(3´,5´)-zyklisches dimerisches Adenosinmonophosphat (c-di-AMP) und dem pegylierten αGalactosylceramid (αGCMPEG) zu dem Polymer-basierten Nanotransporter die frühe Immunantwort derart modifiziert werden, dass die erworbene Immunantwort unterschiedlich polarisiert verläuft.

Die vorliegende Arbeit hat somit nicht nur das Potential der TF-Impfstrategie als Alternative zur klassischen Impfung überhaupt deutlich gemacht, sondern auch dargelegt, dass diese Route bei Verwendung von Nanotransportern in Verbindung mit Adjuvantien ebenfalls für Subunit-Impfstoffe geeignet ist, um z. B. eine schützende Immunantwort gegen ein Influenzavirus zu stimulieren.

3

Introduction

1. Vaccination

a. The ins and outs of vaccination

Infectious diseases represent the second leading cause of death worldwide1. Currently, vaccination is the most effective tool to prevent illness and therefore death in the population. Thus, according to the World Health Organization, vaccines prevent 2 to 3 millions deaths each year and this number could even reach 6 million if all children would receive the recommended vaccine schedule2. Up to now, vaccination allows the control of ten major diseases (i.a. smallpox, diphtheria, tetanus, pertussis, poliomyelitis).

The process of vaccination utilizes the ability of the immune system to generate memory cells specific to encountered pathogens. Therefore, by administrating the whole pathogen or only some specific epitopes, the immune system is able to recognize those foreign molecules (antigens) and build-up a specific humoral and cellular response against these antigens, thereby leading to the protection of the host when confronted with the corresponding pathogen. Edward Jenner is the pioneer of this technique in the Western world. In the 18th century, he inoculated healthy humans with a formulation containing cowpox pustule extract, which has led to the protection of those individuals against the smallpox virus. The birth date of the vaccination is then fixed at the date of the discovery of the protective effect of the cowpox against smallpox: 1796. Later, in 1879, Louis Pasteur scientifically established the general principle of what he named vaccination, in honor of Jenner. Indeed, by studying avian cholera, Pasteur discovered that old bacterial cultures injected to chicken do not kill them but rather protect them against fresh bacteria cultures. He performed similar experiments using rabies virus and finally introduced the new concepts of inactivation, attenuation, and modification through passages3–7.

Although vaccination has already demonstrated its value to protect against infectious diseases, the emergence of new pathogens, as well as the increase of antibiotic resistance, leads to an urgent need of new vaccination strategies. Presently, the vaccine formulations in use are either live-attenuated or inactivated whole pathogens (virus or bacteria) or purified microbial components, the so-called subunit vaccines. The last-mentioned is the most promising formulation candidate for the development of future vaccines. Indeed, containing no living organism and only the subcellular components required to confer protective immunity, their safety profiles not only increase their acceptance in the public

4

but make them especially useful for vaccinating the immunocompromised subpopulation. The use of purified microbial components lower the risk of reactogenicity of the vaccine formulation and therefore makes them more safe. However, the low complexity profile of subunit vaccines usually makes them also less immunogenic, making necessary booster immunizations and/or the inclusion of adjuvants.

b. Challenges for vaccination

Despite the great achievements resulting from vaccination, this field of science is currently facing major challenges. One of them is the logistic constraints in vaccine distribution. Indeed, with the development of new vaccines requiring specific storage conditions, developing countries are facing serious difficulties in their vaccine supply and logistics systems. Another challenge vaccine developers has to face are safety issues, such as the risk of infections and injuries from the use of needles. Likewise, with the increased development of subunit vaccines, a new challenge is rising in the vaccine development field. Indeed, subunit vaccines are composed of a small part of the pathogen of interest, therefore it usually lacks the danger signals needed for activating the immune system and hence are less immunogenic. Thus, this pathogen-derived danger signal should be substituted by an adjuvant. Therefore, new well-defined and effective adjuvants are needed for the development of subunit vaccines8,9. Last but not least, the decline of compliance from the population to get vaccinated is becoming a major challenge as well. It has been estimated that between 5 and 10% of the population has strong anti-vaccination convictions. This leads to the fact that 3 to 7% of all children are under-vaccinated only because of the misgiving of their parents towards some or all vaccines10. It is for all those reasons that the development of stable, safe and painless vaccines is necessary. In this regard, recent approaches aim at the development of non-invasive vaccination strategies targeting mucosal surfaces (e.g. oral, intranasal or transcutaneous vaccination)11–14. The advantages of such vaccination strategies encompass not only the stimulation of efficient local and systemic immune responses already hindering pathogen entry and further dissemination, but also easy and painless administration logistics. Unfortunately, vaccines applied via mucosal surfaces have to overcome some hurdles such as ciliary activity, physical stress (extreme pH) or enzymatic degradation. Thus, part of the antigen will be lost, making necessary increased antigen doses, incorporation of adjuvants and/or carrier systems in order to stimulate efficient immune responses15–17. The same is true in case of

5

transdermal vaccination, when the skin barrier is maintained intact. Thus, also here specific vaccination strategies have to be developed.

2. Skin vaccination

a. Skin anatomy

The skin is the largest organ of the body, it is also known as the integumentary system. Moreover, the skin is the first interface between the body and the external environment, whereby the surface approaches an area of approximately 2 m² in adult humans. This organ has a complex structure composed of different tissues, which enable it to run many functions including the maintenance of the homeostasis. It is constituted of two layers, namely the epidermis, which is directly in contact with the environment and the underlying dermis. Each structure is further subdivided into layers that display specific structures and functions (Fig. 1). Besides protecting the body against external aggression, the skin aims also at maintaining the temperature and prevent the water loss18,19.

Figure 1: Skin anatomy. Adapted from Kabashima et al.20

6

b. Skin Immune System (SIS) and Skin-Associated Lymphoid Tissue (SALT)

Far from being a simple protective barrier, the skin is an organ rich of immune cells that form a dense network of effector cells which actively communicates with the surrounding skin cells. This sophisticated structure has been described by Bos and Kapsenberg as the “skin immune system”21. In the early 80s, Streilein described the immune network of the skin as an analogous tissue to the mucosa-associated lymphoid tissue (MALT), to which he gave the name of skin-associated lymphoid tissue (SALT)22,23. Although MALT and SALT share some common features, they differ regarding functional aspects. Indeed, MALT contains B lymphocytes that form follicles surrounded by T-cells, whereas in the SALT almost all lymphocytes are T-cells which are mainly antigen-experienced memory cells. Nevertheless, the skin does contain immune cells and especially innate myeloid cells that act as antigen-presenting cells (APC) in the context of pathogen invasion. Among them, macrophages and five subpopulations of dendritic cells (DC) have been identified24

(Fig. 2).

Figure 2: Mouse skin dendritic cell subsets. Adapted from Malissen et al24

7

i. Epidermis

The epidermis is the outermost part of stratified epithelium which develops from the ectoderm. Human and mouse epidermis have a similar organization. The human epidermis contains five layers (only four layers in murine skin), namely the stratum corneum (SC) which is the structure in contact with the external environment, the stratum lucidum (present only on the palms and feet soles in human), the stratum granulosum, the stratum spinosum and the stratum basal in contact with the dermis layer (Fig. 1). The epidermis is composed of approximately 95% of keratinocytes, which undergo a maturation process as much as they are moved toward the stratum corneum by the newly formed cells at the basal layer. This maturation process induces modification of the shape of the cell from column-like at the basal layer to flat and dead cells at the stratum corneum. This migration mechanism leads to the shedding of the corneocytes every two weeks25. The epidermis contains also immune cells, for the innate immune compartment, the Langerhans cells which belong to the DC type (Fig. 3).

Langerhans cells (LCs) are radio-resistant cells that establish in the epidermis from yolk sac progenitors before birth and renew themselves independently from bone marrow cells26. It has been shown, in both human and mice, that the maintenance of LCs requires IL-34, which is constitutively produced by keratinocytes27. This mode of renewal used by LCs that is independent of blood-borne precursor cells may constitute an adaptation of the epidermal tissue to the lack of blood vessels of its own. In mice, LCs can be identified by the expression of the MHC class II molecule, CD207, and CD11c. In human, the CD207 is also a specific marker for the LC together with the CD1a24. LC are located above the basal layer of keratinocytes and project their dendrites upward through the tight junctions, which allow them to sample antigens within the cornified layer without breaching the stratum corneum28. More recently, a second population of LC in the epidermis has been identified. This other population is derived from Gr-1hi monocytes which are recruited in the epidermis under inflammatory conditions and have a short life. Therefore, two waves of cells act at the replenishment of the LC in the epidermis: (i) long-term LC-derived from the skin LCs precursors and (ii) short-term LC-derived from blood monocytes. At steady-state, LCs are able to induce the polarization of T-cells into Treg when targeted by self-antigens29. In the context of Candida albicans infection, LCs are migrating toward the draining lymph nodes (LNs) where they secrete high amounts of IL1β, IL-6, and TGF-β, leading to the differentiation and polarization of the LNs naïve T-cells into Th17 cells30. LCs, but also other subsets of dermal DCs, have been shown to induce a B-cell-mediated

8

antibody response by the activation of T-follicular helper cells (Tfh), and especially IgG1 production in the model of Staphylococcus (Fig. 4).

The other immune cells in the epidermis are mainly T-lymphocytes. In the mouse epidermis these lymphocytes are γδ T-cells and display dendritic shape, and they are therefore known as dendritic epidermal T-cells (DETC). Those T-cells express a conserved Vγ5 Vδ1 T-cell receptor (TCR). Upon activation γδ T-cells can produce IFNγ cytokine which induces inflammation and activates other immune cells31. However, in human skin, only a very low number of γδ T-cells is present in the epidermis32.

Figure 3: Human and mouse skin immune cells. From Pasparakis et al.33

ii. Dermis

Unlike the epidermis, the dermis comprises a relatively low amount of cells but consists of about 75% of collagen fibers (percent by dry weight). Another difference between epidermis and dermis is the vascularization. The dermis is highly vascularized, with approximately 60 to 75 capillary loops per millimeter square of human skin. The dermis is also composed of a dense lymphatic capillary network allowing the migration of the immune cells from the skin to the draining LNs25. The murine dermis contains three

9

subtypes of DC namely XCR1+ cDC (with its human equivalent CD141+ cDC), CD11b+

cDC (equivalent to the human CD1c+ cDC) and the double negative cDC (DN cDC) that are the only skin cells expressing a high level of CX3CR1. Although human counterparts have been identified in terms of functionality for most of the murine skin DC, the mouse subtype DN cDC has still no human equivalent. The dermis also contains monocyte-derived DC and macrophages24.

Dermal XCR1+ cDCs are present in both human and mouse dermis at a relatively rare number. Their differentiation requires the factor FLT3L. They express the XC-chemokine receptor 1 (XCR1) as well as the CD207, Clec9a, TLR3, CD141/BDCA3, and the transcription factor IRF8. This population can be subdivided into CD103+ and CD103-

population and Bedoui et al. reported that CD103+ subset is able to cross-present self-antigens as well as skin-associated viral antigens34. They have a high turnover rate and migrate rapidly into the T-cell zone of the draining LNs35. There is an equivalent population in the LNs that expresses the CD8α instead of the CD10336. With this cross-presentation feature, the XCR1+ cDC population is able to induce CD8+ T-cells, particularly in the context of TLR3 costimulation37 (Fig. 4).

CD11b+ cDC is the most abundant DC population in the murine dermis at steady state. This population expresses CD11b, IRF4, CX3CR1, and CCR2. Like the XCR1+ cDCs, the CD11b+ cDCs requires FLT3L for their development38. They predominantly migrate into the paracortex of the draining LNs, in the interfollicular zone39. It has been shown that this DC subset can be activated by the thymic stromal lymphopoietin (TSLP) chemokine which leads to the production of CCL17 and expression of CCR7, and therefore induces T-Helper 2 (Th2) polarization40. Halim et al. showed that the depletion of the group 2 of innate lymphoid cells (ILC2) leads to the impairment of the skin localization of Th2 memory cells upon allergen rechallenge, due to the absence of IL-13 produced by the ILC2, which in turn induces the production of Th2 cell-attracting chemokine CCL17 by the CD11b+

cDCs41 (Fig. 4).

Double negative dermal cDCs (DN cDCs) represent a small population in the dermis which does not express CD207 nor XCR1 but expresses CD11blow, CCR2, and CX3CR1. The development of this population requires FLT3L and depend on IRF4. This population has not been identified in another non-lymphoid or secondary lymphoid tissue. Upon TSLP secretion in the skin, DN cDC population migrates toward the draining LNs and become the most abundant migratory skin DCs population 24-48 h post skin stimulation35. By

10

secreting CCL17 chemokine, this population is believed to target Th2 lymphocyte polarization42 (Fig. 4).

Figure 4: Mouse dendritic cells and their T-cells polarization features. Adapted from Kashem et al42

c. Transfollicular vaccination

Currently, conventional vaccines are delivered using the two main routes of vaccination, namely intramuscular (i.m.) and subcutaneous (s.c.). Those two routes do not aim to target directly resident APC but instead depend on the recruitment of leukocytes that infiltrate the site of vaccination attracted by inflammatory molecules.

Skin tissue allows the use of a large range of vaccine delivery systems including intradermal injection (i.d.) but also needle-free strategies, SC disruptive approaches, and passive techniques such as passive diffusion of protein or DNA as well as transfollicular route43 (Fig. 5). Several animal experiments and clinical trials have already been performed to investigate differents skin vaccine delivery systems (Table 1).

11

Figure 5: Vaccine delivery techniques for cutaneous vaccination. Adapted from44

One of the main obstacles to the development of a non-invasive skin vaccine strategy is the SC barrier. In this regard, the transfollicular vaccination strategy is now being explored as an alternative route for vaccine delivery45–48. Indeed, hair follicles (HF) represent a shortcut across the SC but also allow to deliver the vaccine antigen to the abundant perifollicular APCs. Moreover, the SC is only present in the upper third part of the HF which makes the way through the skin easier49. In addition, Lademann et al. showed that the HF display a storage behavior as applied nanoparticles can be detected inside the HF for up to 10 days50. Other studies, looking at the penetration depth of nanoparticle formulations into the HF, have demonstrated that a diameter of 600 nm allows the deepest penetration into the HF, at around 1400 nm using the porcine skin model51. However, the tight junctions present within the HF form an additional barrier which prevents nanoparticles of penetrating deeper52. Hence, transfollicular (TF) immunization represents a promising route for vaccine antigen delivery, combining skin immune cells targeting, non-invasiveness and painlessness.

12

Table 1: Clinical and pre-clinical skin vaccination (adapted from Combadiere B et al.53)

Technique Immunogens Disruptive technique References

Clinical trials Patch Heat-labile (LT) Lymphotoxin (E. coli)

NO 54

Patch Lymphotoxin E. coli (2 x 10 μg) 2 applications

YES: stratum corneum disruption

55

Patch Rouvax® (Varicella vaccine)

YES: tape stripping before immunization

56

Patch Lymphotoxin E. coli (2 x 35.5 μg) 2 applications

YES: abrasive technique 57

Transcutaneous anti-influenza vaccine Agrippal (season 2004–2005)

YES: cyanoacrylate skin surface stripping (CSSS)

58

Pre-clinical experimentations

Patch Cholera Toxin (CT) YES: short resting time after shaving

59

Topical application

Inactivated HSV-1 + CT

Yes: no resting time of the skin before immunization

60

Patch Adenovector C tetanus toxin

- 61

Patch Yersinia pestisPeptides + CT

YES: immunization directly after shaving

62

Patch HIV-1 Peptides + CT or Lymphotoxin

YES: stratum corneum abrasion

63

Patch + oleic or retinoic acid

Inactivated Influenza virus + CT

YES: stratum corneum abrasion

64

Naked skin + immiquimod®

OVA protein YES: tape stripping before immunization

65

Patch + Lipid C +/- CT/CPG

Chlamydia external membrane protein in lipid C

YES: acetone before immunization

66

13

d. Targeting of the skin dendritic cells

As mention in section 2.b, each skin DC is able to drive the adaptive immune response toward a specific polarization of the T-lymphocytes. Therefore, targeting the vaccine antigen to a distinct subset of DC would allow orientating the immune response in the needed direction according to the pathogen. For example, the cross-presenting XCR1+ cDC subset is of interest to target when an antiviral Th1 immune response is required. There are different ways to target the skin DCs in cutaneous vaccination. Special targeting could be obtained by administering the antigen directly to the location of the specific DC. However, this method usually requires breaking the skin barrier in order to access the DCs of interest. Skin DCs can also be targeted by customizing the chemical and physical properties of the vaccine antigen. Indeed, the detection and uptake of a pathogen or antigen are critical steps for the development of effective adaptive immune responses. Therefore, the size, shape, and molecular structure of the antigen can be customized to target a specific DC67. It has been shown that particles with a size between 20 and 200 nm are better uptaked by the DCs (human or murine) as compared to bigger or smaller particles, and with a higher uptake observed for 100 nm. Also, in vivo, polymeric particles in the size of 200 nm encompassing an antigen are longer presented by APCs (72 h) compared to those with a size of 40 nm (48 h)68–71.

3. Vaccine formulation for transdermal application

a. Use of inactivated viruses as vaccine antigens

The goal of vaccination is to generate a strong immune response toward a specific antigen which in turn is able to induce long-term protection against the corresponding pathogen. Inactivated virus vaccines are usually made using chemical or physical agents in order to remove the infectivity while retaining the immunogenicity of the virus. Thus, β-propiolactone is already used for some human rabies vaccines and has the advantage of not damaging the protein antigen. However, other agents used to inactivate viruses may affect the immunogenicity of the antigen resulting in modified and thus, less efficient immune responses (e.g. shortening the immune response, narrowing the viral antigen spectrum, weakening the mucosal immune responses). This is, for example, the case when formalin is used as inactivation agent, where many antigens are irreversibly changed.

14

b. Nanoparticles as non-viral vehicles

Conventionally, vaccine formulations contain killed pathogens (inactivated), attenuated viruses or subunit protein antigens. Although very effective, live-attenuated vaccines suffer from safety issues due to their risk for individuals who are immunocompromised as a result of disease or therapeutic interventions. On the other hand, inactivated pathogens, as well as subunit vaccines, induce weaker and sometimes incomplete immune responses. In order to compensate these weaknesses incorporation of an adjuvant is required. With the development of nanomaterials, micro- and nanoparticle-based delivery systems are now offering a unique possibility to develop safe and effective vaccines able to replace the traditional viral vectors. Indeed, those biomaterials are made of lipids, polymers or polysaccharides which leads to their biocompatibility and biodegradable behaviors. For example, chitosan, PLGA (polylactic-co-glycolic acid), and polyphosphazene (PCPP) polymers are “generally regarded as safe” (GRAS) by the Food and Drug Administration (FDA) or even already used for biomedical application in humans72. Thus, nanoparticles (NPs) have already been widely studied for their use as vaccine delivery system and/or adjuvant effects73. Moreover, NPs can be used for specific cell targeting, indeed by decorating the nanoparticle surface with a cellular ligand, the NP can selectively release its payload in the cell of interest74,75. Compared to soluble antigen alone, the delivery of an antigen using nanoparticles can result in an about 30-fold increased uptake by DCs76,77. In addition, the composition of the nanoparticles (e.g. charges, biomolecules used, etc.) directly affects the interaction with the target cells in terms of cell entry and subsequent stimulation of adaptive immune responses. For example, antigens entering by the endocytic pathway preferentially stimulate Th2 responses, whereas entry by membrane fusion usually results in Th1 dominated immune responses78. Nowadays, the most studied polymeric nanoparticles are based on PLGA and PLA (polylactic acid). The class of polyphosphazene polymers has also been intensively studied for their potential role in drug delivery systems. They also display immunostimulatory effects, making them useful as adjuvants. Vaccine formulations containing PCPP polymer have been reported as safe79. Polyphosphazenes are water-soluble, high molecular weight synthetic polymers. The two most studied members of this family are the PCPP (poly[di(carboxylatophenoxy)phosphazene]) and the PCEP (poly[di(sodiumcarboxylatoethylphenoxy)phosphazene]). They have demonstrated their capacity to enhance the magnitude, quality and duration of the immune response when co-formulated with vaccine antigens80–87. Moreover, polyphosphazene polymers have already been used to enhance the immune response in influenza vaccines88–90. Likewise, the water-

15

soluble properties of those carrier systems give to those macromolecules the potential to be dissolved and to release the antigen in a hydrated environment, such as the skin. Studies have even highlighted the potential of the PCPP and PCEP to enhance humoral and cellular immune responses following vaccination via the skin route88–92.

c. Adjuvants

Adjuvants are molecules used to enhance the immune response toward the vaccine antigen. This concept comes from Ramon et al, in the 1920’, who observed that horses that develop an abscess at the injection site of diphtheria toxoid showed a higher specific antibody titer93,94. Later, in 1926, Glenny et al demonstrated the adjuvant effect of aluminum compounds95. It is in 1936 that Freund develops an emulsion containing killed mycobacteria which results in the most potent adjuvant known, the so-called “Freund’s complete adjuvant” (FCA)96,97. Adjuvants can be used for different purposes: (i) enhancement of the immunogenicity of a certain antigen; (ii) reduction of the amount of antigen or the number of immunizations needed; (iii) increase of immune cell targeting and uptake98. In addition, adjuvants can be also used to tailor the antigen-specific immune responses according to clinical needs (Table 2). Finally, the route of vaccine application is also a factor that should be taken into account when choosing the adjuvant (parenteral adjuvants, mucosal adjuvants, …). In this regard, it is unlikely that a single adjuvant is able to achieve all the properties that are needed in all the future vaccines. Therefore, different adjuvants are needed, depending on the pathogen, to stimulate the immune response able to confer protection against this specific pathogen. More recently, some studies have addressed the possibility to use a combination of adjuvants in order to improve vaccine efficacy99,100.

16

Adjuvant Immune Stimulatory Component Mechanism of Action Immune

Response Indication

Alum Aluminum salts Release of host DNA from dying cells and

proteasome activation Antibodies, Th2

e.g., HAV, HBV, HPV, Haemophilusinfluenza, tetanus,

diphtheria

MF59, AS03

Oil–in-water (squalene in water) emulsion

Mechanism unknown, enhanced uptake by

APCs

Antibodies, Th1/Th2, long-lived memory

Pandemic influenzastrains (H5N1,

H1N1)

MLP Non-toxic derivative of LPS TLR 4 agonist Ab, Th1 HBV, HPV

AS04 MPL plus alum Activation of NF-kB pathway Ab, Th1 HBV, HPV

CTB Cholera toxin B subunit

Binding to gangliosides provide T-cell co-stimulation

Ab (mucosal IgA) Cholera (orally)

VLP, Virosomes

Self-assembling viral proteins

PAMP signals, TLR-like APC activation

Ab, Th1/Th2, memory HBV, HPV

Liposomes (IRIV)

Proteoliposomes composedof phospholipids,

influenza hemagglutinin

HA induced improved uptake by APCs Ab, Th1/Th2 HAV, influenza

DETOX MPL plus mycobacterial cell wall skeleton

Tumor antigen induced T-cell stimulation Th1 Cancer

BCG Bacillus Calmette– Guérin

Promote cell-mediated immune reactivity to

tumor-associated antigens Th1 Cancer

HAV: Hepatitis A virus ; HBV: Hepatitis B virus ; HPV: Human papillomavirus ; MLP: Monophosphoryl lipid A ; VLP: Virus-like particles

Table 2: Adjuvants licensed for human use (adapted from Riese et al.15)

d. Influenza virus

Influenza A viruses (IAV) belong to the Orthomyxovirus family. IAV is composed of a lipid envelope, which is taken from the host cell where the virus multiplied, containing three transmembrane proteins known as hemagglutinin (HA), neuraminidase (NA) and an ion channel formed by the matrix 2 protein (M2). Beneath this envelope is located the viral protein M1, this protein forms a shell and gives the rigidity to the virus. This protein core

17

contains eight segments of negative-sense, single-stranded RNA coding for 11 proteins which include, in addition to the previously cited proteins, the nucleoprotein, non-structural proteins 1 and 2 (NS1, NS2), polymerase acidic protein (PA) and polymerase basic proteins 1 and 2 (PB1, PB1-F2 and PB2)101. Influenza viruses are subdivided by the antigenic variation of their HA and NA proteins. Currently, there are 18 HA and 11 NA serotypes known.

When the virus enters the host through the oral or nasal cavities, it reaches and attaches to the respiratory epithelium via the HA protein. After penetrating the cell, the influenza virus starts to spread within the respiratory tract. Nevertheless, the viral RNA in the infected cells is recognized by the TLR3 and TLR7 resulting in the secretion of proinflammatory chemokines and cytokines including type I interferons (IFNs). The TLR family is express on a multitude of innate immune cells aiming at the host defense against pathogens. TLR molecules are composed of homologous proteins with an extracellular domain which contains a leucine-rich repeat and an intracellular which mediate the signal transduction. Upon activation, they induce the expression of proinflammatory cytokines102–105. This local inflammation mediates the recruitment of immune cells including neutrophils, monocytes and NK cells, for the clearance of the virus106,107. After this first line of innate immune cells recruitment, the adaptive arm of the immune system will take over with the generation of CD4+ and CD8+ T-cells, as well as B-cells for the clearance of the viruses, elimination of infected cells, and generation of an immune memory108.

Influenza A viruses are responsible for global health burden leading to up to 650,000 deaths every year due to the respiratory diseases linked to seasonal flu infection109. Furthermore, a study based on the United States population estimates the total annual cost impact of the influenza epidemic at an average of $81.1 billion110. Therefore, vaccination is the most cost-efficient prevention method to reduce the influenza infection by controlling the spreading of the virus among the population. Until now, three classes of seasonal influenza vaccines are licensed: (i) inactivated; (ii) live-attenuated and (iii) recombinant HA protein. The inactivated influenza virus (IIV) is injected via i.m. route, usually composed of two influenza A subtypes together with an influenza B subtype, it is therefore called trivalent vaccine (TIV). Recently, a quadrivalent vaccine (QIV) has been licensed which include a second influenza B subtype in addition to the TIV. This class of vaccine induces a predominant IgG antibody response toward the HA protein of the virus. The live-attenuated vaccines contain a similar mix of influenza subtypes than the QIV but are administrated via the intranasal route and therefore stimulate, in addition to the systemic immune

18

responses, a mucosal immune response. In that case, IgA antibodies and cytotoxic T-lymphocytes (CTL) are major actors in the neutralization and clearance of the virus13,111. Although this class of vaccine is very efficient to induce protective immunity, their use in high-risk populations (i.e. young children, elderly, or immunocompromised individuals) is not recommended, mainly because of a high risk of a severe systemic reaction against the vaccine112. Finally, the last group of licensed influenza vaccines is composed of four recombinant HA proteins (FluBlok). Clinical trials have shown that in children younger than 3 years Flublok induces a lower hemagglutinin inhibition titer (HI) compared to the licensed influenza vaccine for this population. Also, the lack of safety and effectiveness studies in the populations between 3 to 18 years old and older than 65 years leads to the exclusion of those populations for the use of the Flublok. Therefore, this vaccine is approved for adults between 18 to 49 years old who are allergic to eggs113.

All in all, licensed seasonal influenza vaccines have a mean efficacy at about 60% in healthy adults and 83% in children, respectively114,115. However, the vaccine strains and the circulating epidemic virus often do not match completely and thus, only suboptimal protection of the population is obtained. Moreover, on the one hand the inactivated influenza vaccine does not stimulate robust immune responses in elderly (<20% protection)116 and on the other hand, the live-attenuated vaccine is not licensed for the population above 49 years, putting a vulnerable section of the population under poor protection117. Ideally, the influenza virus must protect all age groups, be safe and manufactured without eggs, as well as overcome the lack of patient compliance and protect against all influenza A viruses: the so-called “universal influenza vaccine”.

19

Aim of the thesis

The development of non-invasive needle-free vaccination strategies for subunit vaccines which are able to stimulate protective immune responses against emerging and reemerging pathogens is a critical need in public health.

Therefore, the overall aim of this thesis was to develop a non-invasive TF vaccination strategy, based on nanoparticle formulations able to protect against influenza virus infection. In order to achieve this goal, we first proofed if vaccination via the TF route can elicit protective immunity. To this end, mice were immunized with inactivated influenza A virus H1N1 or H5N1, which are currently been used for human vaccination. After validating the TF route with the inactivated virus, we moved on toward subunit vaccine formulation and investigated if vaccines based on polymeric nanoparticles co-formulated with the adjuvants c-di-AMP or αGCMPEG are able to stimulate protective immune responses against influenza infection. In this regard, we performed studies to gain a better understanding of the initiation of the antigen-specific immune responses stimulated after TF vaccination. Using recombinant influenza HA protein incorporated in polyphosphazene polymers in a similar prime-boost vaccination strategy, the early innate immune events occurring in the skin after TF vaccination were unraveled. The resulting knowledge should facilitate the development of new vaccination strategies which are safe and confer effective protection against infectious diseases.

20

Materials

1. Instruments

Table 3: Instruments and consumables

Instruments Company

Cell counter, Z2 Coulter counter Beckman Coulter, Germany

Cell culture plates Greiner, Germany

Cell Harvester ICH-110 INOTECH, Switzerland

Cell strainer 100 μm BD Bioscience, USA

Cell tric 50 μm PArtec GmbH, Germany

CTL Immunospotanalyser Cellular Technology, Ltd., Germany

ELISA Reader Synergy2 BioTek, Germany

Falcon tubes Greiner, Germany

Flow cytometry tubes 1.2 ml MP Biomedicals, France

Flow cytometry tubes 5 ml BD Bioscience, USA

Flow cytometer Fortessa BD Bioscience, USA

Flow cytometer AriaII Sorter BD, USA

High binding ELISA plates Greiner, Germany

Incubator, Heracell 240i Thermo Scientific, Germany

Leica SP5 upright Leica Microsystems, Germany

Pipettes (0.2–2 μl, 2-20 μl, 20-200 μl, 100-1000 μl) Thermo Scientific, Germany

Pipettboy Thermo Scientific, Germany

Plate washer ELx 405 BioTek, Germany

21

Precision 2000 automated pipetting system BioTek, Germany

Scintillation Counter Wallac 1450 PerkinElmer, Finland

Serological pipets Roth, Germany

Sterile hood Hera Safe Thermo Scientific, Germany

Thermomixer compact Eppendorf, Germany

Two-Photon Microscopy Zeiss LSM710 Zeiss, Germany

Vortex Genie-2 Omnilab, Germany

2. Chemical compounds and reagents

Table 4: Chemical compounds and reagents

Chemicals Company 2,2’-azino-bis (3-ethyibenzthiazoline-6 sulfonic acid) diammonium salt (ABTS) Sigma-Aldrich, Germany

3-amino-9-ethyl-carbazole (AEC substrate) BD Pharmingen, USA

Acetic acid (CH3COOH) Merk, Germany

Ammonium thiocyanate Carl-Roth, Germany

Ampuwa® Serumwerk, Germany

Avidin-HRP (horseradish peroxidase) conjugated BD Pharmingen

Bovine serum albumin (BSA) Sigma-Chemie, Germany

Brefeldin A Sigma-Chemie, Germany

Carboxyfluorescein diacetate succinimidyl ester (CFDA-SE) Invitrogen, Germany

Collagenase D Roche, Germany

Concanavalin A from canavaliaensiformis Sigma-Aldrich, Germany

Dimethyl sulfoxide (DMSO) Sigma-Aldrich, Germany

22

Dispase II Protease from Bacillus polymyxa Sigma-Aldrich, Germany

DNAse Roche, Germany

Ethylenediaminetetraactic acid (EDTA) Fluka, Switzerland

Ethanol 100% Fluka, Switzerland

Fetal calf serum, South America (FCS) Greiner Bio-One, USA

Formaldehyde > 36.5% Riesel-de-Haën, Germany

Gentamicine Gibco, UK

L-Glutamine Gibco, UK

HEPES Gibco, UK

Hydrogen peroxide (H2O2) Sigma-Aldrich, Germany

[methyl-3H]Thymidine 5.0 Ci / mmol Amersham, Bioscience

Ionomycin BD, USA

Isofluran® Curamed vet inhalation anesthetic Essex Tierarznei, Germany

2-Mercaptoethanol (50 mM) Gibco, UK Penicillin/streptomycin (100 units/ml penicillin G sodium; 50 μg/ml streptomycin sulfate in 85% saline) Gibco, UK

Phorbol 12-myrisate 13-acetate (PMA) Sigma-Aldrich, Germany

RPMI 1640 medium (+L-glutamine) Gibco, UK

Receptor destroying enzyme (RDE) Sigma-Aldrich, Germany

Sodium acetate (CH3COONa) Merk, Germany

Sodium carbonate (Na2CO3) Carl-Roth, Germany

Sodium chloride (NaCl) Carl-Roth, Germany

Sodium dihydrogen phosphate (NaH2PO4) Merk, Germany

Sodium hydrogen carbonate (NaHCO3) Merk, Germany

Sodium hydroxide (NAOH) Carl-Roth, Germany

Sulfuric acid Sigma-Aldrich, Germany

23

Tris/HCl Sigma-Aldrich, Germany

Trypsin Sigma-Aldrich, Germany

Tween 20 Carl-Roth, Germany

3. Antibodies

Table 5: Antibodies used for flow cytometry

Antigen Fluorochrome Clone Dilution Company

CD3 V500 500A2 1:200 BD Horizon, USA

CD3 PE-Cy7 145-2C11 1:100 BD Horizon, USA

CD4 PE-Cy7 RM4.5 1:1000 eBioscience, Germany

CD8 BV650 53-6.7 1:300 BioLegend, USA

CD11b BV605 M1/70 1:300 BioLegend, USA

CD11c BV785 N418 1:200 BioLegend, USA

CD11c FITC N418 1:500 eBioscience, Germany

CD19 BV785 6D5 1:800 BioLegend, USA

CD24 APC M1/69 1:300 BioLegend, USA

CD44 BV785 1M7 1:1000 BioLegend, USA

CD45 PerCP-Cy5.5 30-F11 1:1000 eBioscience, Germany CD45R (B220) PE-Cy7 RA3-6B2 1:600 BD Horizon, USA

CD62L BV605 MEL-14 1:250 BioLegend, USA

CD64 BV711 X54-5/7.1 1:200 BioLegend, USA

CD69 PE H1.2F3 1:200 BD Horizon, USA

CD86 BV650 GL1 1:200 BioLegend, USA

CD103 BV510 2E7 1:600 BioLegend, USA

24

IFNγ BV711 XMG1.2 1:300 BioLegend, USA

IFNγ BV785 XMG1.2 1:200 BioLegend, USA

IL-2 APC-Cy7 JES6-5H4 1:200 BD Horizon, USA

IL-4 APC 11B11 1:200 eBioscience, Germany

IL-10 BV421 JES5-16E3 1:150 BioLegend, USA

IL-17 BV605 TC11-18H10.1 1:200 BioLegend, USA

IL-17 FITC eBio17B7 1:500 eBioscience, Germany

Ly6G PE-Cy7 1A8 1:600 BioLegend, USA

MHC-II BV421 M5/114-15-2 1:250 BioLegend, USA

NK1.1 PE-Cy7 PK136 1:1000 eBioscience, Germany

TNFα PerCPeF710 MPG-XT22 1:200 eBioscience, Germany

γδTCR PE GL3 1:200 BioLegend, USA

Table 6: Antibodies used for Enzyme-Linked Immunospot Assay (ELISpot)

Antibody Conjugate Dilution Company Anti-mouse IFNγ Purified 1:200 BD Pharmagen, USA

Anti-mouse IFNγ Biotinylated 1:250 BD Pharmagen, USA

Anti-mouse IL-2 Purified 1:200 BD Pharmagen, USA

Anti-mouse IL-2 Biotinylated 1:250 BD Pharmagen, USA

Anti-mouse IL-4 Purified 1:200 BD Pharmagen, USA

Anti-mouse IL-4 Biotinylated 1:250 BD Pharmagen, USA

Anti-mouse IL-17A Purified 1:200 eBioscience, USA

Anti-mouse IL-17A Biotinylated 1:250 eBioscience, USA

Streptavidin Horseradish Peroxydase (HRP) 1:100 BD Pharmagen, USA

25

Table 7: Antibodies used for Enzyme-Linked Immunosorbent Assay (ELISA)

Antibody Conjugate Dilution Company

Goat, anti-mouse IgA Biotinylated 1:5000 Southern Biotech, USA

Goat, anti-mouse IgG Biotinylated 1:5000 Sigma-Aldrich, Germany

Goat, anti-mouse IgG1 Biotinylated 1:5000 Southern Biotech, USA

Goat, anti-mouse IgG2a Biotinylated 1:5000 Southern Biotech, USA

Goat, anti-mouse IgG2c Biotinylated 1:5000 Southern Biotech, USA

Goat, anti-mouse IgM Biotinylated 1:200 Sigma-Aldrich, Germany

Streptavidin Horseradish Peroxydase (HRP) 1:1000 BD Pharmagen, USA

4. Solutions, buffers and culture media

Table 8: Solutions, buffers and culture media

Solution/Buffer Composition ABTS (2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) 0.3 g/L ABTS in 0.1 M citric acid

ABTS + H2O2 0.03% (v/v) H2O2 in ABTS solution

Acetate solution (0.1 M) 148 ml acetic acid, 0.2 mM + 352 ml of 0.2 mM sodium acetate; adjust volume to 1L with H2O, pH 5.0

Aminoethylcarbazole (AEC) stock solution 100mg AEC substrate in 10 ml DMF

AEC + H2O2333.3 μl AEC stock solution in 10 ml 0.1 M acetate solution, 5μl H2O2 (30%)

Blocking Buffer (ELISA) PBS + 3% BSA

Carbonate buffer (0.1 M) 4.401 g NaHCO3 + 1 L H2O, pH 8.2

26

Calcium saline solution pH 7.2 1 g CaCl2 • 2H2O ; 9 g NaCl ; 1.2 g H3BO3 ; 0.052 g Na2B4O7 • 10H2O in 1 L of H2O

Dilution buffer (ELISA) PBS + 0.1% Tween 20 + 1% BSA

Dilution buffer (ELISPOT) 1x PBS containing 10% FCS

1 x PBS (phosphate-buffered saline) 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4*7H2O, 0.24 g KH2PO4 in 1L H2O, pH 7.2

Permeabilization buffer 0.5% BSA + 0.5% saponin/PBS (w/v)

Receptor Destroying Enzyme solution 1 vial of RDE (Sigma) in 5 ml Milli-Q water + 95 ml calcium saline solution, pH 7.2

RPMI complete 10% v/v FCS heat inactivated (h.i.), 100 U/ml penicillin, 50 μg/ml streptomycin, 2 mM L-glutamine

Skin cells wash buffer 1 X PBS + 2mM EDTA + 2% FCS

Skin digestion buffer RPMI 1640 + 2% FCS + 3.2% 1M HEPES + 1% Pen/Strp

Wash buffer (ELISA) PBS + 0.1% Tween 20

Wash buffer (ELISPOT) PBS + 0.05% Tween 20

5. Kit for cytokine detection and cell enrichment

Table 9: Kits for cytokine detection and cell enrichment

Kits Company LEGENDplex Th Cytokines (13-plex) Biolegend, USA

LEGENDplex Mouse proinflammatory Chemokines (13-plex) Biolegend, USA

MagniSort Mouse CD4+ T-cell enrichment kit Invitrogen, USA

MagniSort Mouse CD8+ T-cell enrichment kit Invitrogen, USA

MagniSort Mouse T-cell enrichment kit Invitrogen, USA

27

6. Antigens, adjuvants and nanocarriers

Table 10: Antigens, adjuvants and nanocarriers

Antigens/adjuvants/nanocarrier Company

-galactosylceramide polyethylene glycol (αGCMPEG) HZI, Germany

Chitosan/PLGA nanoparticles HIPS, Germany

Bis-(3'-5')-cyclic dimeric-adenosine-monophosphate (c-di-AMP) BioLog, Germany

Bisphosphorothioate analog of 2'3'-c-di-AMP, (Rp;Rp) cyclic [A(2’,5’)psA(3’,5’)ps] Invivogen, USA

H1N1 antigen (NIBSC 13/164) NIBSC, UK

H5N1 (NIBRG-14; see 4.1.10) NIBSC, UK

Polyphosphazene polymer (PCEP) VIDO, Canada

Polyphosphazene polymer (PCPP) VIDO, Canada

Recombinant HA1 protein HZI, Germany

Ovalbumin (OVA; Grade VII, LPS free) Invivogen, USA

FITC-labeled WH peptid Creative peptides, USA

7. Influenza virus

Table 11: Influenza strains used for animal challenge

Strain Description

PR8 H1N1 Ova SINFEKL CD8 Recombinant H1N1 A/PR/8/34 which contains the OVA epitope SIINFEKL (OT-I PR8) was offered by Dr. D Topham (University of Rochester Medical Center)

H1N1 PR8 Freiburg Mouse adapted H1N1 A/PR/8/34 provided by Dr. P Blazejewska and Dr. K Schughart (Helhholtz Center for Infectious Research)

H5N1 Reassortant prepared by reverse genetics fromA/Vietnam/1194/2004 (H5N1) virus and A/PR/8/34(H1N1) virus (NIBSC, UK)

28

Influenza Luciferase Influenza A/PR8/33 PASTN (H1N1)

Influenza A/PR8/33 PASTN (H1N1), expressing the nano luciferase protein downstream of the PA polymerase subunit sequence

29

Methods

1. Immunization protocol

Female C57BL/6 mice 6 to 8 weeks of age were immunized by TF route, two days after shaving of an area of approximately 2 cm² using first a trimer and then VEET© depilation cream. The vaccinated area was covered with tape for 24 h. When vaccinating using protein antigen, mice received a total of 3 to 4 immunizations every two weeks on skin of the back. For kinetic experiments, mice received only one immunization on the inner side of the ears and without previous depilation. When using inactivated influenza virus, the mice were immunized only three times. In all this study, when adjuvanted formulation was used, 10 µg of adjuvant per dose was applied on the skin and 7.5µg/dose was used for the s.c. and i.n. routes.

2. Sample collection and processing

In order to study at a single cell level the outcome of the immunization protocols, mice were euthanized using CO2 after the retro-orbital complex bleeding. Blood samples were centrifuged for 5 min at 8000 x g and sera were stored at -20°C for further analysis. Spleen and skin-draining lymph nodes were collected on ice in complete RPMI medium. Skin tissue (back skin or ears) was collected on a PBS solution containing dispase II.

a. Spleen

Single-cell suspension of splenocytes was obtained by mashing the tissue on a 100 µm cell mesh using the extremity of a 2 ml syringe plunger. The cell solution was transferred into a 15 ml falcon tube and centrifuged at 300 x g for 8 min at 4°C. The supernatant was discarded and for lysis of the erythrocytes, the pellet was resuspended with 1 ml of ACK lysis buffer for 1 min at room temperature (RT). The lysing process was stopped by adding 10 ml of complete RPMI. After samples were centrifuged at 300 x g for 10 min at 4°C the supernatant was discarded and the pellet resuspended in complete RPMI prior determination of the cell concentration.

30

b. Skin-draining lymph nodes

To investigate the adaptive immune responses stimulated in the tissue draining the vaccination area, cervical (when immunized on the ears), brachial, axillary and inguinal (when immunized on the back) skin-draining lymph nodes (LNs) were isolated and transferred into complete RPMI medium. As it was the case for the spleen, LNs were mashed onto a 100 µm mesh using syringe plunger. The single-cell solution was collected in a 15 ml falcon tube and centrifuged at 300 x g for 10 min at 4°C. Next, cells were washed using complete RPMI medium (or PBS if the cells were directly used for flow cytometry straining) and filtered before the determination of the cell concentration.

c. Skin tissue

After euthanasia of mice, ears were collected, washed with 70% ethanol, and the dorsal and ventral side were separated with the help of two forceps. Each half was incubated dermis side down on a dispase II solution [0.4 mg/ml] (Roche) for 2.5 h at 37°C. The skin was then minced in small pieces and transferred in a solution of 2 mg/ml of collagenase D plus 0.5 mg/ml of DNase in medium (RPMI + 2% FCS + 1% PenStrep + 3.2% 1M HEPES) and incubated for 45 min at 37°C under agitation. Skin digestion solution was then filtrated through a 50 µm cell strainer and washed in PBS + 2 mM EDTA + 2% FCS before determination of the cell concentration and antibody-based staining flow cytometric analysis.

3. Determination of cell concentrations

An aliquot of the cell suspension was diluted in an isotonic solution at a ratio of 1:1000. Cell numbers were determined using the Multi32 Coulter Z2® (Beckman Coulter) with the multi-size analyzer software. Only cells with a diameter between 5-10 µm were counted in order to exclude cell debris and tissue aggregates.

31

4. Detection of cytokine-producing cells using the ELISpot assay

The number of antigen-specific cytokine-producing cells was determined using an ELISpot assay. To this end, 96-well plates (BD Pharmingen) were coated with anti-IFN-γ, anti-IL2, anti-IL4 or anti-IL17 antibodies overnight at 4°C. Then, plates were washed one time with culture medium (RPMI, 10% fetal calf serum (FCS), PenStrep, L-glutamine, and β-mercaptoethanol) and blocked with the same medium for 1 h at RT. Cells were seeded in culture medium at a concentration of 5x105 splenocytes/ml in the presence or absence of the corresponding antigen (10 µg/ml of protein antigen (OVA or HA), 2 µg/ml of OVA peptide SIINFEKL and 0.5 µg of HA equivalent/ml (HAeq/ml) of inactivated influenza virus). The SIINFEKL peptide is specific for the MHC-I molecule which will activate specifically the CD8+ T-cells, where OVA protein will mainly activate the CD4+ T-cells. Plates were incubated 24 h for IFN-γ and 48 h for IL-2, IL-4, and IL-17. Then, cells were removed and the plates processed according to the manufacturer’s instructions. Colored spots were counted with an ELISpot reader (CTL-Europe GmbH) and analyzed using the ImmunoSpot image analyzer software v3.2.

5. Antigen-specific multifunctional T-cells analysis

To investigate the functionality and polyfunctionality of antigen-specific T-lymphocytes stimulated after vaccination, splenocytes were seeded at the concentration of 1x107 cells/ml in 6 well plates and incubated 18 h at 37°C in the presence or absence of the vaccine antigen (10 µg/ml for OVA and HA proteins and 0.5 µg/ml for H1N1 and H5N1 antigen). After 18 h, brefeldin A (5 µg/ml) and monensin (6 µg/ml) were added to the culture and cells were incubated for further 6 h in order to block the extracellular secretion of cytokines. Finally, cells were collected in FACS tubes, washed in PBS and stained for surface marker and intracellular cytokines, as described below in section 8.

6. Cell proliferation assay

The ability of spleen-derived immune cells to proliferate upon restimulation with the vaccine antigen as well as their cytokine profile was measured 96 h post restimulation. The proliferative capacity was determined by the incorporation of 3H-thymidine. Splenocytes from vaccinated animals were incubated in 96-well plates at the concentration of 5x105

32

cells/ml (flat-bottom for proliferation and U-bottom for cytokines) with increasing concentrations of antigen for 96 h in quadruplicates at 37°C 5% CO2 and 95% humidity. For analyzing proliferation, 18 h before harvesting the cells, 1 µCi of 3H-thymidine was added to each well. Then, cells were harvested on a filter paper (Filtermat A; Wallac, Freiburg, Germany) by using a cell harvester (Inotech, Wohlen, Switzerland). Melt-on scintillator sheets were melted on the filter allowing measuring beta irradiation using a scintillation counter (Wallac 1450, Micro-Trilux). The number of proliferating cells was indirectly determined by counting the beta radiation emitted by [3H]-thymidine incorporated into the DNA of proliferating cells. Results are expressed as the ratio of values from stimulated and non-stimulated samples [stimulation index (SI)].

7. Multiplex FlowCytomix (Cytometric Bead Array (CBA))

The presence of mouse Th1/Th2/Th9/Th17 cytokines and/or proinflammatory chemokines was determined using a FlowCytomix immunoassay according to the manufacturer's instructions (Biolegend). The cytokine profile was measured in the supernatant of antigen-restimulated splenocytes derived from immunized mice as previously described for the proliferation assay.

When analyzing the concentration of proinflammatory chemokines in skin tissue homogenates and sera, the proinflammatory chemokine panel of Biolegend was used.

8. Cell staining for flow cytometric analysis

Surface staining

Single-cell suspensions from spleen, LNs or skin were transferred into FACS tubes or V-bottom plates and centrifuged at 300 x g for 5 min at 4°C. After been washed with PBS, samples were centrifuged again at 300 x g for 5 min at 4°C and the cell pellets were resuspended in 100 µl PBS containing antibodies specific for the surface markers of choice as well as the fixable live/dead dye. Cells were stained for 30 min at 4°C in the dark. For staining of skin-derived cells, samples were beforehand incubated with αCD16/αCD32 antibodies to block the Fc receptors in order to decrease unspecific staining. In addition, all incubation and washing steps were performed in PBS + 2 mM EDTA + 2% FCS (skin

33

cells buffer). In figures 6 and 7 is displayed the gating strategy for identification of the four subsets of skin dendritic cells (Fig. 6) and skin migrating cells in the LNs (Fig. 7). After incubation, cells were washed two times with PBS (or skin cells buffer). If no intracellular staining was needed, cells were directly acquired using the BD Fortessa flow cytometer.

Figure 6: Gating strategy used to distinguish the different murine skin DC subsets. Gating on the live/Dead negative population allows excluding the dead cells. Doublet cells were excluded in two steps using first FSC-A vs FSC-H and then SSC-A vs SSC-W. Immune cells were gated as CD45+ and lineage positive (CD3, CD45R, NK1.1, and Ly6G) cells were excluded. Dendritic cells were identified as MHC-II positive cells and the four DCs subsets were gated using CD24 vs CD11b.

34

Figure 7: Gating strategy used to identify the murine skin migratory DCs in the draining LNs. Gating on the live/Dead negative population allows excluding the dead cells. Doublet cells were excluded in two steps using first FSC-A vs FSC-H and then SSC-A vs SSC-W. Cells positive for CD3, CD45R, NK1.1, and Ly6G were identified as lineage positive and therefore excluded. Then, migratory DCs were identified as MHC-IIhi and CD11c positive.

Intracellular staining

For intracellular staining, directly after the last washing step following the surface staining, cells were resuspended in 100 µl of the Cytofix/Cytoperm buffer (BD Bioscience) and incubated for 25 min at 4°C in the dark. After this step, the cells were washed using the Perm/Wash buffer solution and centrifuged at 300 x g for 5 min at 4°C. Then the cells were resuspended in Perm/Wash buffer containing the intracellular antibody mix and incubated for 30 min at 4°C in the dark. Afterward, samples were washed with Perm/Wash buffer and resuspended in PBS for direct flow cytometer acquisition.

Single staining and compensation beads were prepared for all used fluorochromes and used for the analysis using FlowJo software to assign every dye to its proper spectrum.

35

9. Antigen-specific antibody detection and functionality assessment

a. IgG and IgG subclasses ELISA

Antigen-specific antibody titers in sera of immunized mice were determined using ELISA. High binding protein plates were coated with 2 µg/ml of OVA protein or recombinant HA protein, or with H1N1 or H5N1 virus (1 µg HAeq/ml) in 0.05 M carbonate buffer overnight at 4°C. After blocking unspecific binding sites using 3% BSA in PBS, a serial 2-fold dilutions of sera in 3% BSA/PBS were added to the plate (100 μl/well). For the plates coated with HA protein, the blocking buffer used was TBS-T + 5% skim milk. The same buffer was used for samples dilution and incubation. After 1.5 h incubation at 37°C, plates were washed six times using 200 µl/well of 1% BSA/PBS/0.05% Tween 20 (or 1% skim milk in TBS-T buffer for the HA protein ELISA) and the secondary antibodies biotinylated goat anti-mouse IgG, IgG1, IgG2a or IgG2c, respectively, from Sigma ( USA) were added. After 1 h incubation at 37°C, plates were washed six times as previously and samples were incubated for 30 min at 37°C in the presence of peroxidase-conjugated streptavidin (BD Pharmingen, USA). Finally, after another six washing steps, reactions were developed using ABTS [2, 20-azino-bis(3- ethylbenzthiazoline-6-sulfonic acid)] in 0.1 M citrate-phosphate buffer (pH 4.35) containing 0.1% H2O2. Endpoint titers are expressed as absolute values of the last dilution giving an optical density (OD405 nm) being two times higher than the values of the negative control after 15 min incubation, as previously described118.

b. Hemagglutination Inhibition assay (HAI)

i. RDE treatment of sera

Serum of immunized mice was diluted at the ratio 1:5 in RDE solution and incubated overnight at 37°C. Then, the diluted serum was further diluted at a ratio 1:1 in 1.5% sodium citrate and heated up at 56°C for 30 min to inactivate the RDE.

ii. Standardization of the Red Blood Cells (RBC)

Chicken red blood cells were centrifuged at 150 x g for 10 min at 4°C. Then, the supernatant was removed and the pellet washed two times with 2 volumes of 0.9% NaCl

36

solution. After washing, the RBC pellet was resuspended in 2 volumes of 0.9% NaCl solution, the cells were counted and the concentration was adjusted to 4,000 cells/ml, corresponding to a theoretical 1% of RBC solution. This cell concentration was used for the HAI assay.

iii. HAI test for sera

A serial two-fold dilution of each treated sera was prepared in a V-bottom plate with a final volume of 25 µl for each well. The antigen was titrated in order to contain 4 HA units/25 µl = standardized antigen. HA unit is the amount of virus needed to agglutinate an equal volume of red blood cells. A volume of 25 µl of standardized antigen (H1N1 of H5N1) was added to all the wells. The plates were gently shaken and incubated for 30-45 min at RT. After the incubation step, 50 µl of standardized RBC solution was added and plates were further incubated for 30 min at RT. The HAI titer was finally recorded as the highest serum dilution where no complete hemagglutination of the RBC is observed.

c. Enzyme-Linked Lectin Assay (ELLA)

As for the HAI, RDE treated sera samples were diluted following a serial twofold dilution and transferred in a 96 well plate (50 µl per well) coated with Fetuin. Then, 50 µl of the antigen (H1N1 or H5N1) was added to each well and incubated between 16 to 18 h at 37°C in a humidified incubator. Plates were washed six times with PBS-tween, prior addition of 100 µl per well of peanut agglutinin. Then, plates were incubated for further 2 h at RT in the dark. Afterward, plates were washed three times with PBS-tween and 100 µl of OPD in citrate buffer was added for 10 min at RT in the dark. The reaction was then stopped by adding 100 µl of 1N H2SO4. Finally, the OD was measured at 490 nm. The 50% NA-inhibition antibody titer was determined as the highest dilution that resulted in at least 50% inhibition of the maximal signal.

10.Naïve T-lymphocytes magnetic enrichment

A negative selection of spleen and lymph node derived T-cells was performed using the MagniSort Mouse CD4+ T-cell enrichment kit (Invitrogen). According to the kit

37

instructions for CD4+ T-cell enrichment, cell suspensions were incubated with anti-mouse biotinylated antibodies specific for CD8, CD11b, CD19, CD24, CD45R, CD49b, Ly-6G, γδ TCR and TER-119 for 10 min at RT. After a washing step using PBS, streptavidin-conjugated magnetic beads were added to the cell suspension. By positioning the samples then in a magnetic field, labeled cells stuck to the resin while non-labeled cells were recovered. After another washing step, cells have been used for in vivo and in vitroexperiments. The same negative enrichment procedure has been followed for CD8+ T-cells using the MagniSort Mouse CD8+ T-cells enrichment kit (Invitrogen).

11.Adoptive transfer

To analyze antigen-specific CD4+ and CD8+ T-cell activation after a single immunization, naïve T-cells from OVA-MHC-II-, OVA-MHC-I- or HA-TCR-specific transgenic mice were labeled using carboxyfluorescein succinimidyl ester (CFSE), a green fluorescent dye which is able to bind covalently intracellular proteins. The labeled cells were then injected intravenously (i.v.). CFSE is used as a tracking dye as well as a tool to quantify the proliferation of the cells. Indeed, with every cell division, all fluorescent molecules are equally distributed among the daughter cells, reducing the fluorescent intensity detected by the flow cytometer by 50%. Thus, when analyzing the proliferative capacity of the cells, the CFSE signal is displayed as a histogram where every peak from high intensity to low intensity represent one division step.

12.In vivo determination of cytotoxic T-lymphocytes (CTL)

For detection of antigen-specific cytotoxic T-lymphocytes, splenocytes from wild type (WT) mice were divided. One half was pulsed with the CD8-peptide of OVA protein for 2 h at 37°C in complete RPMI, the other half was cultured in the absence of an antigen peptide. Afterward, the cells were washed two times in PBS and resuspended in PBS containing CFSE dye. The pulsed cells were incubated with 5 µM of CFSE whereas the non-pulsed cells were incubated with a concentration of 0.5 µM of CFSE for 5 min at RT. After incubation, an equal volume of FCS was added to both group of cells and incubated for 5 min at RT to stop the reaction. Finally, cells were washed two times in PBS and pulsed and unpulsed cells were pooled at a ratio 1:1 before to be injected i.v. to vaccinated mice. After 48 h, spleens, LNs, and lungs of transplanted mice were collected, processed

38

for single-cell suspension and analyzed by FACS. The killing capacity of antigen-specific CD8+ T-cells was then assessed as a correlate of reduced CFSE positive cell frequency, according to the following formula:

% ����� = 100 − �% �� ���� ����� ���

% �� ������� ��� ���� �������� �� % �� ���� ����� ���

% �� ������� ��� ���� ����� �� � � 100

13.Challenge of vaccinated mice

Mice vaccinated with OVA were challenged using 3 x 102 foci forming unit (ffu) of an H1N1/PR8 strain that expresses the OVA SIINFEKL peptide diluted in PBS (20 µl) via the intranasal route. Mice vaccinated with inactivated H5N1 virus were challenged using the mouse-adapted H5N1 virus at the dose of 2 x 103 ffu. Mice vaccinated with HA protein were challenged using the H1N1/PR8/PASTN strain expressing the luciferase reporter gene. Mice healthy status (weight and behavior) was monitored for the next following 14 days. Where indicated, mice were sacrificed and considered as killed by the infection according to a score approved by the local government.

14.In vivo Imaging System (IVIS)

The in vivo imaging was performed using the IVIS 2000 system. Mice were injected i.v. in the lateral vein of the tail with the enzyme’s substrate solution NanoGlo reagent in PBS at a dilution of 1:20. Then mice were anesthetized with isoflurane and image acquisition was performed with Living Image software (PerkinElmer). Images were analyzed using the same software than for the acquisition

15.Skin explant culture

After euthanasia, ears of mice were collected, soaked for 1 min with 70% ethanol, and left until processing in PBS solution on ice. Then, the dorsal and ventral side were separated

39

with the help of two forceps. The ventral halves (were the vaccine formulation was applied) were cultured in 24-well plates (one per well) for 72 h in complete culture medium: RPMI (Gibco, Life technology) containing 10% FCS, 2 mM L-glutamine (Gibco, Life technologies), 50 µg/ml gentamicin (Gibco, Life technologies) and 50 µM β-mercaptoethanol (Gibco, Life technologies). Emigrating cells were then collected in the bottom of the wells, washed with PBS, stained with fluorochrome-labeled antibodies and analyzed by flow cytometry.

16.In vivo imaging using 2-Photon Microscopy

For the in vivo imaging, mice were immunized on the inner side of the ear as already described in the first paragraph of the methods section. After 4 h, mice were anesthetized and placed on a heating stage adjusted to 37°C. The outer side of the ear was fixed to a metal platform and a coverslip was sealed on the top of the ear. The two-photon imaging was performed using the Zeiss LSM 700 system with a Mai Tai DeepSee Ti:Sa laser (Spectra-Physics) tuned at 920 nm. The emitted signal was split by 625 nm, 490 nm, and 520 nm long pass and dichroic mirrors and filtered with 640-710 nm, 509-522 nm, 500-550 nm, and 570-610 nm filters. The second-harmonic generation (SHG) corresponds of the interaction of two photons with the same frequency with a nonlinear material (in our case the collagen fibers of the skin) and generate a new photon with half of the wavelength. Images were analyzed using the Fiji software (HIH).

17.Statistical analysis

Data were analyzed using GraphPad Prism 8 software. Statistical significance of the observed differences were analyzed using a student’s t-test for the comparison of two independent groups and two-way ANOVA test for the comparison of multiple groups. Differences were considered significant at p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****) respectively.

40

Results

1. Vaccination strategy

a. Skin integrity before immunization

The aim of the present work was the development of a non-invasive nanoparticle-based vaccination strategy targeting the dermal immune system. To this end, animal work has been performed based on different mouse models. In order to have full access to the skin for immunization and to avoid trapping of the vaccine formulations by electrostatic interactions with the hair, the skin has been shaved 48 h before vaccination. As shown by Mittal et al. only after 48 h resting time the skin integrity reaches the basal level again as indicated by the relative water loss of the skin119. This is of interest since we aimed in developing a non-invasive vaccination strategy avoiding skin disruption prior immunization which would increase the risk of cross-infections. Thus, in a first attempt recruitment of neutrophils into the skin, reflecting inflammatory reactions has also been studied following depilation. In line with the results obtained by Mittal et al. the steady-state is restored after 48 h (Fig. 8). Therefore, all the following experiments in this project have been performed using this shaving/immunization timeline.

0 1 24 48 120 1680

1

2

3

4

5

Time post depilation (hours)

Figure 8: Skin integrity post depilation indicated by neutrophil influx into the skin following depilation of murine skin.

41

b. No oral “cross-vaccination” by grooming

In order to keep the vaccine formulation on the skin and avoid any oral “cross-vaccination” by licking while grooming, the vaccinated area was always protected for 24 h using a special medical tape. Cross-vaccination has been investigated by co-housing skin vaccinated mice with non-vaccinated animals for 7 days following vaccination. In order to increase the detection of “cross-vaccination”, CFSE labeled naïve CD4+ T-cells isolated from HA-TCR transgenic mice were injected intravenously one day prior to the immunization. The TCR of the injected CD4+ and CD8+ T-cells is specific for the immunodominant MHC-II and MHC-I HA peptide, respectively. Activation of those cells by APC was evaluated measuring their proliferative capacity by FACS, as indicated by the reduction of the CFSE signal. None of the non-vaccinated mice showed cell proliferation neither at local (draining LN-derived T-cells) nor at systemic level (spleen-derived T-cells). In contrast, LN- and spleen-derived CD4+ and CD8+ T-cells of vaccinated mice showed a strong proliferative capacity (Fig. 9).

Figure 9: No oral cross-vaccination by grooming following transfollicular vaccination. HA-specific naive T-cells were labeled and adoptively transfered into WT mice before a single immunization. Data represent the percentage of proliferating CFSE labeled cells derived from (A) draining LNs and (B) spleen. Standard error of the mean (SEM) is indicated by vertical lines, (n=3). Differences are statistically significant with P<0.05 (*) and P>0.01 (**) compared to immunized Co-Housed group).

42

2. Proof of concept: Efficient non-invasive transfollicular delivery of inactivated influenza virus as a vaccine antigen

a. Adoptive transfer model

Recent studies carried out by Mittal et al.45,119 have shown the ability of the TF route to stimulate antigen-specific adaptive immune responses when using OVA-loaded nanoparticles (NP). Although antigen-specific humoral and cellular immune responses have been induced using this route, no protection upon pathogen challenge has been described so far. Therefore, we decided to investigate whether the TF vaccination route is able to induce a protective immune response when using the inactivated influenza virus. Considering the similar size of the influenza virus compared to the NP used in the study of Mittal et al. (180 nm), it appeared to be the perfect choice to investigate TF vaccine delivery. First of all, in order to prove whether the vaccine antigen applied by TF route is delivered to the skin APC, uptaked, processed and presented by them to antigen-specific T-lymphocytes, we performed a proof-of-concept study based on an adoptive transfer model. Thus, the immune system of C57BL/6 mice was enriched by i.v. injection on HA-specific T-lymphocytes 24 h before the immunization with inactivated H1N1 virus equivalent to 5 µg of HA adjuvanted with the STING agonist c-di-AMP. Spleen and draining LNs were sampled 7 days post-immunization. Both, in the draining LNs and in the spleen of immunized mice, we were able to determine the proliferation of antigen-specific T-lymphocytes. Thus, this adjuvanted formulation was able to stimulate CD4+ as well as CD8+ T-lymphocytes, as indicated by up to 80% proliferating cells in the LNs and around 70% in the spleen (Fig. 10).

43

Figure 10: HA-specific proliferation of CD4+ and CD8+ lymphocytes by c-di-AMP adjuvanted inactivated H1N1 virus. HA-specific naive T-cells were labeled and adoptively transferred into WT mice before a single immunization. Data represent the percentage of proliferating labeled cells (A) in the draining LNs and (B) in the spleen. Standard error of the mean (SEM) is indicated by horizontal lines, (n=3). Differences are statistically significant with P<0.05 (*) and P>0.01 (**) compared to the control group.

b. Characterization of the adaptive immune responses stimulated following vaccination with non-adjuvanted inactivated H1N1.

The results obtained using the adoptive transfer model provided the proof-of-concept for the planned vaccination strategy. However, in order to investigate more in detail the immune responses stimulated following TF immunization, a vaccination experiment has been performed in naïve mice and humoral and cellular responses have been investigated. The used vaccination schedule is based on previous publications as well as on in-house studies119,120. The adaptive immune response has been characterized two weeks after the last immunization (Fig. 11A). In line with what has been previously published119,120, TF vaccination of mice with the non-adjuvanted H1N1 formulation stimulated reasonable antigen-specific IgG titers (Fig. 11B). On the other hand, the observed IgG1/IgG2c subclass ratio of about 100 indicates a T helper 2 (Th2) dominated immune response (Fig. 11C).

44

Figure 11: Systemic humoral immune responses induced in mice after three immunizations with inactivated H1N1 virus via TF route. (A) Vaccination schedule. (B) H1N1-specific IgG titers in sera 13 days after the last immunization. (C) H1N1-specific IgG1 and IgG2c titers in the sera of immunized mice. Results are displayed as average of the last serum dilution (endpoint dilution, e.p.d) showing the double value (OD405nm) of the control background. SEM is indicated by vertical lines, (n=12). Differences are statistically significant with P>0.01 (**) with respect to the control group.

Beside humoral immune responses, also the type of cellular response stimulated plays a critical role in efficiently combating a pathogen once an infection occurs. Thus, when immunizing mice with the inactivated H1N1 virus, the proliferative capacity of the antigen-specific splenocytes upon antigen restimulation is reaching twice the baseline (i.e. non-vaccinated animals) (Fig. 12A). Moreover, the maximal proliferation rate is already reached using an antigen concentration of no more than 0.5 µg of HAeq/mL, indicating an efficient stimulation of the immune system following TF vaccination using inactivated H1N1 (Fig. 12A). In order to decipher which lymphocyte subset is stimulated by

Prime 1st Boost 2nd Boost Samling

-2 14 28 420 12 26

Shave Shave Shave

Vaccination scheduleA

45

vaccination with the inactivated H1N1, the proliferative capacity of cells derived from draining LNs has been monitored using the CFSE dilution assay. Despite the fact that the differences were not statistically significant, there is a clear trend for increased numbers of CD8+ T-lymphocytes and B-lymphocytes, with seven and three times higher values than in the control group, respectively (Fig. 12B).

0.0 0.5 1.0 1.5 2.0

Figure 12: Cellular immune response stimulated by non-adjuvanted inactivated H1N1 virus following TF vaccination. (A) Splenocytes from vaccinated animals were restimulated for 96 h with different concentrations of inactivated H1N1 virus (starting from 0.1µg HAeq/ml to 2 µg HAeq/ml). The proliferative capacity of the cells was determined by incorporation of 3H-thymidine. Results are the average of quadruplicates and are expressed as SI. Statistically significant differences when comparing the control group with respect values in mice receiving the inactivated H1N1 antigen were assessed using two-way ANOVA with Dunnett’s multiple comparison test, with p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) and p < 0.0001 (****). (B) Draining LNs derived cells from vaccinated animals were stained with CFSE dye and restimulated with inactivated H1N1 virus (0.5 µg HAeq/ml). Data are expressed as number of antigen-specific proliferating cells per 500000 LNs cells. The proliferative capacity of the cells was assessed after 96 h by FACS. SEM is indicated by vertical lines, (n=3).

For further characterization of the cellular responses, we investigated the cytokine profiles stimulated with the inactivated virus in the speen after TF vaccination. As shown in Fig. 13, immunization with the inactivated influenza virus H1N1 stimulated the production of Th1 cytokines such as IFNγ, IL-2, and TNFα, and Th2 cytokines like IL-4, IL-6, and IL-10, as well as the Th17 cytokine IL-17. The obtained results indicate a mixed Th1/Th2 response further confirming the observed IgG1/IgG2 ratio.

46

Figure 13: (A) The number of cytokine-producing cells (IL-2, IL-4, and IL17) was determined by ELISpot. Results are expressed as spot-forming units of 5x105 restimulated cells. Background was removed using unstimulated samples. SEM is indicated by vertical lines (n=3). (B) Cytokines secreted by splenocytes of vaccinated mice (pg/ml) cultured for 96 h in the presence of the vaccine antigen (0.5 µg HAeq/ml).

Next, the protective capacity of the immune responses stimulated following TF immunization with inactivated H1N1 has been investigated by infecting mice with a sub-lethal dose of the influenza strain H1N1 three weeks after the last immunization. Mice of the positive control (i.e. vaccinated with the inactivated H1N1 virus via s.c. route), as well as mice vaccinated via TF route did not display any weight loss post-infection, thereby indicating full protection against influenza infection. In contrast, none-vaccinated mice showed a drop reaching a maximum of about 10% of weight loss at day 8 post-challenge (Fig. 14).

47

Figure 14: Protection of vaccinated animals against influenza A virus infection. C57BL/6 mice (n=6) received three immunizations with non-adjuvanted H1N1 virus by TF route. Three weeks after the last immunization, mice were challenged with a sub-lethal dose (2x102 ffu) of the influenza strain A/PR8/F-1 (H1N1) given by i.n. route (10 µl per nostril). The animal body weight was monitored for a period of 14 days. Results are expressed as average percentage of body loss.

c. Tailored immune response using adjuvanted formulations.

Since vaccination with inactivated H1N1 virus stimulated protective immune responses against virus infection in mice, the next step was to investigate if the addition of an adjuvant would allow a reduction of the antigen dose and/or the number of immunizations. Two different adjuvants have been used for this purpose: c-di-AMP and αGalCerMPEG

The c-di-AMP, a secondary bacterial messenger which signals for DNA integrity, is involved in biofilm formation and regulates bacterial virulence and mobility121,122. C-di-AMP activates the signaling molecule STING (stimulator of interferon genes) which leads to the activation of TBK3-IRF3 pathway resulting in the production of type I interferons and TNFα123. The second adjuvant, αGalCerMPEG (αGCMPEG), is a pegylated derivative of the well-known α-galactosylceramide, which is originally derived from the marine sponges Agelas mauritianus. This glycolipid was shown to be presented by CD1d molecules on APC and therefore act as a ligand for invariant Vα14+ NKT-cells, which in turn secrete upon activation large amounts of IFNγ and IL-4124,125. In order to investigate the possible reduction of antigen dose and/or immunization numbers, we first attempt to

48

compare the strength and quality of the humoral and cellular immune responses in the same vaccination settings described above. As shown in figure 15A, co-formulation of αGCMPEG with the inactivated H1N1 virus leads to significantly higher systemic antibody titers as compared to the control group than the co-formulation with c-di-AMP or immunization with H1N1 alone. However, no significant differences have been observed between adjuvanted and non-adjuvanted groups (Fig. 15A). Interestingly, while co-administration of inactivated H1N1 with αGCMPEG resulted only in a minor modification of the direction of the stimulated immune response, co-administration of c-di-AMP redirects the H1N1-specific immune responses to a more balanced Th1/Th2 response (Fig. 15B and C). Thus, c-di-AMP acts on the balance between IgG1 and IgG2c by raising the IgG2c antibody titers resulting in an IgG1/IgG2c ratio of around 3, which is 16 and 8 times lower as compared to those obtained for H1N1 alone or co-administered with αGCMPEG, respectively (Fig. 15B and C). Taken together, TF vaccination with H1N1 alone drives the immune responses toward a Th2 dominated response, whereas by adding c-di-AMP a more balanced Th1/Th2 response was obtained. In contrast, the modulation by αGCMPEG is only marginal, if any at all.

Control

H1N1 A

ntigen

H1N1 A

ntigen

+ c-d

i-AMP

H1N1 an

tigen

+GCMPEG

H1N1 a

ntigen

H1N1 an

tigen

+ c-di-A

MP

H1N1 a

ntigen

+ GCMPEG 1

10

100

1000

10000

100000

1000000 IgG1IgG2c

H1N1 an

tigen

H1N1 a

ntigen

+ c-di-A

MP

H1N1 a

ntigen

+ GCMPEG

A B C

Figure 15: Systemic humoral immune responses induced in mice after three immunizations with adjuvanted H1N1 virus antigen via TF route. (A) H1N1-specific IgG titers in sera 13 days after the last immunization. Each bullet symbol indicates an individual animal. (B) H1N1-specific IgG1 and IgG2c in the sera of immunized mice. Results are displayed as average of the last sera dilution (end point dilution, e.p.d) showing the double value (OD405nm) of the control background. (C) Antibody titer ratio IgG1/IgG2c. SEM is indicated by vertical lines. Differences are statistically significant with P<0.05 (*) and P<0.001 (**) compared to control group.

49

Next, we investigated the neutralizing activity of the antibodies produced after TF immunization toward the influenza virus using two different assays, the hemagglutination inhibition (HAI) assay which allows detecting the antibodies that can neutralize the virus by targeting the hemagglutinin protein located on the virus envelop and the Enzyme-Linked Lectin Assay (ELLA) allowing detecting neutralizing antibodies specific for another envelope protein of the virus, the neuraminidase. Remarkably, while mice vaccinated with inactivated H1N1 alone showed very different levels of neutralizing antibody titers, mice receiving inactivated H1N1 co-administered with c-di-AMP produced almost identical antibody titers against HA and NA (Fig. 16). Although less striking compared to c-di-AMP, also co-administration of αGCMPEG stimulated equal levels of antigen-specific antibody production among mice within the group (Fig. 16). This suggest that the quality of the stimulated antibody responses is more robust and considerably enhanced by adjuvantation.

Figure 16: Anti-HA and anti-NA neutralizing antibody titers in sera of vaccinated mice were measured by determining the HAI titers (black bar, n=3) for the anti-HA antibodies and by ELLA for anti-NA antibodies. For the HAI, results corresponds to the highest dilution of sera that prevent complete hemagglutination of 4 HA units of hemagglutination/25 µl. For the ELLA, data represent the highest dilution of sera that results in at least 50% inhibition of the maximal signal.

50

In the next step, the effect of the adjuvants on the stimulated cellular immune responses both at local and systemic level has been analyzed as described in section 2b. TF vaccination of mice with inactivated H1N1 co-administered with αGCMPEG stimulated stronger cellular proliferation as compared to those observed immunizing mice with virus alone or co-administered with c-di-AMP (Fig. 17). Likewise to what was already observed when mice were immunized using inactivated H1N1 virus alone, the two main populations activated by the adjuvanted formulations were CD8+ T-cells and B-lymphocytes (Fig. 17B). However, adjuvantation resulted in up to threefold increased levels of CD4+ T-cells, up to sevenfold increased numbers of CD8+ T-cells and up to fourfold increased numbers of antigen-specific B-cells (Fig. 17B).

Stim

ulat

ion

inde

x (S

I)

Control

inactiv

ated H1N

1

inactiv

ated H1N

1 + c-

di-AMP

inactiv

ated H1N

1 + GCMPEG

Num

ber o

f ant

igen

-spe

cific

pro

lifer

atin

g ce

lls

A B

Figure 17: Cellular immune responses stimulated by adjuvanted inactivated H1N1 virus. (A) Splenocytes from vaccinated animals were restimulated with increasing concentrations of the inactivated H1N1 virus (starting from 0.1 µg HAeq/ml to 2 µg HAeq/ml). Proliferation of the cells following 96 h of incubation was determined by incorporation of 3H-thymidine. Results are the average of quadruplicates and are expressed as SI. Statistically significant differences compared to the control group with respect to H1N1 antigen (*), H1N1 antigen + c-di-AMP (o) or H1N1 antigen + αGCMPEG (+) were determined using two-way ANOVA with Dunnett’s multiple comparison test, with p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) and p < 0.0001 (****). (B) Lymphocytes derived from draining LNs of vaccinated animals were stained with the CFSE dye and restimulated with 1 µg HAeq/ml of inactivated H1N1 virus. The proliferative capacity of the cells was assessed after 96 h by FACS. Data are expressed as number of antigen-specific proliferating cells per 500000 LNs cells. Draining LNs cells from mice belonging to the same group (n=3) were pooled.

51

When evaluating the number of cytokine-producing cells stimulated after vaccination, both adjuvanted formulations resulted in increased cell numbers as compared to those detected using the antigen alone (Fig. 18A). In the same line, when analyzing the profiles of secreted cytokines after vaccination, both adjuvanted formulations showed increased concentrations of IFNγ, IL-2, IL-6, IL-10, TNFα, and IL-17A as compared to those detected using the antigen alone. (Fig. 18B). Moreover, only when adding an adjuvant remarkably increased IFNγ and IL-2 titers have been stimulated with respect to those obtained immunizing with H1N1 alone (Fig. 18B). Nevertheless, both adjuvants stimulated a similar cytokine profile, whereby in line with previous studies c-di-AMP also stimulated the production of the Th17 cytokines IL-17A and IL-22 (Fig. 18B).

0 20 40 60 80

H1N1 antigen + GCMPEG

H1N1 antigen + c-di-AMP

H1N1 antigen

Control

H1N1-specific cytokine producing cells(sfu/5.105 cells)

IL-2 IL-4 IL-17

*

***

*

**

IFN-g

A B

Figure 18: (A) The number of cytokine-producing cells (IL-2, IL-4, IL-17) was determined by ELISpot. Results are expressed as spot-forming units of 5x105 restimulated cells. Background values of unrestimulated samples were subtracted. SEM is indicated by vertical lines, (n=3). Differences are statistically significant with P<0.05 (*) and P>0.01 (**) compared to control group. (B) Cytokines secreted by splenocytes of vaccinated mice (pg/ml) cultured for 96 h in the presence of the vaccine antigen (0.5 µg HAeq/ml).

d. Adjuvanted H5N1 formulation improve strength and quality of the immune response even when using low antigen doses

From the previous experiment, we learned that co-administration of the adjuvants c-di-AMP and αGCMPEG results in the stimulation of stronger antigen-specific cellular responses and modulate the quality of the elicited humoral immune responses, thereby leading to production of antibodies with increased biological activity. Thus, in order to

52

validate the possibility of antigen dose sparing as well as to investigate whether similar results will be obtained using a different antigen, we repeated the experiment described above using a 2.5 times lower dose of another influenza antigen, the inactivated H5N1 influenza virus. Using the same readout, we have been able to characterize and to compare the strength and the quality of the immune responses stimulated after TF vaccination using 2 µg of HA equivalent of inactivated H5N1 influenza virus with or without co-administration of an adjuvant. In addition, we evaluated the potential of a c-di-AMP derivative which due to sulfonation is protected against enzymatic degradation by phosphodiesterases126: 2’3’-c-di-AM(PS)2. This molecule persists longer in the vaccinated animals, which could in turn result in a protracted immune stimulation and thus an improved immune response.

Investigating the systemic humoral immune response, we observed that inactivated H5N1 alone stimulates lower IgG-specific antibody titers (1:2583) as compared to the formulations adjuvanted with c-di-AMP (1:15033) and 2’3’-c-di-AM(PS)2 (1:27733), respectively. These titers are comparable with those obtained using 2.5 times higher concentrations of inactivated H1N1 virus (Fig. 15). However, only co-administration with 2’3’-c-di-AM(PS)2 seems to result in the production of statistically higher antibody titers as compared to the control group (Fig 19.A). In line with the results obtained using H1N1 antigen, co-administration of c-di-AMP or its derivative modified the stimulated immune responses towards a more balanced Th1/Th2 response compared to the Th2 dominated response stimulated by the antigen alone, as indicated by the four (c-di-AMP) and two (2’3’-c-di-AM(PS)2) times lower IgG1/IgG2c ratios (Fig. 19B and C).

53

H5N1

-spe

cific

IgG

tite

r in

sera

(e.p

.d)

Control

H5N1 a

lone

H5N1 +

c-di-A

MP

H5N1 +

2'3'-

c-di-A

M(PS)2

Ratio

IgG

1/Ig

G2c

H5N1 a

lone

H5N1 +

c-di-A

MP

H5N1 +

2'3'-

c-di-A

M(PS)2

H5N1 a

lone

H5N1 +

c-di-A

MP

H5N1 + 2'

3'-c-d

i-AM(P

S)2

A B C

Figure 19: Systemic humoral immune responses induced in mice after three immunizations with adjuvanted H5N1 virus antigen via TF route. (A) H5N1-specific IgG titers in sera 13 days after the last immunization. Each bullet symbol indicates an invidual animal. (B) H5N1-specific IgG1 and IgG2c in the sera of immunized mice. Results are displayed as average of the last dilution of sera (end point dilution, e.p.d) showing the double value (OD405nm) of the control background. (C) IgG1/IgG2c antibody titer ratio. SEM is indicated by vertical lines. Differences are statistically significant with P<0.05 (*) compared to control group.

While no relevant differences in the humoral responses stimulated by 2’3’c-di-AM(PS)2 and c-di-AMP could be observed, the combination of c-di-AMP with the H5N1 antigen seems to be slightly more efficient at stimulating cellular response, as indicated by the significantly increased proliferative capacity of splenocytes from vaccinated mice restimulated with 2 µg/ml of HAeq H5N1 virus as compared to those observed in mice receiving H5N1 with 2’3’c-di-AM(PS)2 (Fig. 20A). Although not statistically significant, c-di-AMP seems to be more efficient at stimulating CD8+ and CD4+ T-cells with respect to 2’3’c-di-AM(PS)2, as indicated by increased numbers of proliferating cells. In contrast, the proliferative capacity of B cells does not seem to be significantly affected by c-di-AMP or its derivative (Fig. 20B).

54

Stim

ulat

ion

inde

x (S

I)

Control

H5N1 alo

ne

H5N1 + c-d

i-AMP

H5N1 + 2'3

'-c-di-A

M(PS)2

A B

Figure 20: Cellular immune responses stimulated by the adjuvanted inactivated H5N1 virus. (A) Splenocytes from vaccinated animals were restimulated with increasing concentrations of inactivated H5N1 virus (starting from 0.1µg HAeq/ml to 2 µg HAeq/ml). After 96 h, the proliferation of the cells was determined by the incorporation of 3H-thymidine. Results are the average of quadruplicates and are expressed as SI. Statistically significant differences for H5N1 alone, H5N1 + c-di-AMP or H5N1 + 2’3’-c-di-AM(PS)2 as compared to control group were evaluated using two-way ANOVA with Dunnett’s multiple comparison test, with P < 0.05 (*), P < 0.01 (**), P < 0.001 (***) and P < 0.0001 (****). (B) Draining LNs from vaccinated animals were stained with the CFSE dye and restimulated with 0.5µg HAeq/ml of inactivated H1N1 virus. Data are expressed as number of antigen-specific proliferating cells per 500000 LNs cells. The proliferative capacity of the cells was assessed after 96 h by FACS. SEM is indicated by vertical lines, (n=3). Differences are statistically significant with P<0.05 (*), P<0.01 (**), P<0.001 (***) and P<0.0001 (****) compared to control group.

Evaluation of the cytokine profiles stimulated following TF vaccination with inactivated H5N1 alone or co-administered with an adjuvant does not showed any significant differences between c-di-AMP and its derivative 2’3’ c-di-AM(PS)2 (Fig. 21). Remarkably, while the number of cytokine-producing cells has not been significantly increased by incorporation of either adjuvant as compared to antigen alone, both c-di-AMP and 2’3’ c-di-AM(PS)2 stimulated increased levels of cytokines (e.g. IFNγ and IL-17) secreted by the cells (Fig 21.A and B). Furthermore, beside IFNγ, H5N1 antigen alone is only able to induce the production of low levels of the T helper cytokines IL-4, IL-6, and IL-10 upon antigen restimulation. On the other hand, when adjuvanted with c-di-AMP, not only the expression of those three cytokines further increases, but also those of the Th 17 cytokines IL-17A, and IL-22 (Fig. 21B). All in all, the observed cytokine profiles are in

55

line with the obtained IgG subtype ratios, thereby indicating the stimulation of mixed Th1/Th2/Th17 responses by the two c-di-nucleotides.

Besides analyzing the stimulated cytokine profiles we additionally evaluated the quality of the stimulated H5N1-specific T-cell responses by investigating their capacity to produce more than only the effector cytokine. In fact, it has been shown that the presence of multifunctional T-cells are associated with enhanced protection against certain infections127–131. When investigating the polyfunctionality of the antigen-specific CD4+ T-lymphocytes obtained from spleens of vaccinated mice, 2’3’ c-di-AM(PS)2 adjuvanted H5N1 shows a significantly higher frequency of multifunctional T-cells compared to H5N1 alone (Fig. 21C). No significant differences have been detected between the two c-di-AMP molecules. In addition, although not statistically significant, co-administration of c-di-AMP and its derivative seems to further increase the quality of the stimulated cellular response, as indicated by enhanced frequencies of tri-functional CD4+ T-cells (i.e.production of IFNγ+, TNFα+, and IL-2+) (Fig. 21C).

56

Figure 21: (A) The number of cytokine-producing cells (IFNγ, IL-2, IL-4, and IL17) was determined by ELISpot. Results are expressed as spot-forming units of 1x106 restimulated cells. (B) Heatmap showing the cytokine pattern of splenocyte culture supernatants after 96 h of antigen restimulation (1 µg HAeq/ml). (C) The histogram displays the frequency of multifunctional CD4+ T-lymphocytes in the spleen 24 h after in vitro antigen stimulation and determined by flow cytometry analysis. The pie charts represent the percentage of each bi- (IFNγ/IL-2, IFNγ/TNFα, IL-2/ TNFα) or tri-functional (IFNγ/IL-2/TNFα) CD4+ T-cells among the total multifunctional cells. Background was removed using unrestimulated samples. SEM is indicated by vertical lines, (n=3). Differences are statistically significant with P<0.05 (*) and P<0.01 (**) compared to control group).

57

3. Can polymeric nanoparticles be utilized for the transfollicular vaccination route: the polyphosphazene polymer

So far we demonstrated that TF immunization in principal can stimulate protective immune responses against infection. We also demonstrated that the antigen dose can be reduced when adjuvants are incorporated into the vaccine formulation. Nevertheless, complex and costly production processes like for the influenza virus make attractive the development of subunit vaccines based on synthetic antigens instead of using inactivated virus. In this regard, recent studies revealed already the necessity of using NP for the delivery of vaccine antigens via the TF route allowing entering the HF and subsequently being taken up by the resident immune cells resulting in antigen-specific immune responses45. In addition, those NP protect the vaccine cargo against degradation. However, often the described approaches require coarse epilation in order to increase vaccine efficacy and/or the used NPs fail to stimulate all arms of the immune system132,133. Therefore, new NPs need to be developed allowing targeting of specific skin immune cells in order to stimulate broader antigen-specific immune responses. To this end, we investigated the usefulness of the polyphosphazene polymer family, especially PCPP and PCEP for TF vaccination. Indeed, previous studies have already confirmed the usefulness of those polymers to protect the vaccine cargo when applied by i.n. route as well as to stimulate the immune cells by displaying an adjuvant effect134,135. Polyphosphazenes are also currently studied for their use in combination with microneedles-based vaccination systems92. In this section, we are investigating the innate and adaptive immune responses following TF immunization using OVA as a model vaccine antigen.

a. The nanocarrier penetrate into the hair follicle

Previous studies have demonstrated that the use of nanoparticle for skin vaccination allows delivering of antigens into the hair follicle45. In order to confirm that the polyphosphazene formulations applied on the skin also reaches the hair follicles, we used the so-called 2-photon microscopy which allowed us to follow the fluorescent-labeled OVA loaded in PCPP. The figure 22A shows two hair follicles of a mouse ear skin without any formulation applied to it. In figure 22B, we can see the red fluorescent OVA protein penetrating inside the hair follicle along with the hair (autofluorescent in green). The figure

58

22C represents a scanning through the skin layers from the dermis (number one) to the epidermis (number 10) in 4 µm steps.

Figure 22: In vivo 2-photon microscopy imaging. Penetration of OVA protein labeled with the AF647 dye (red) and loaded in PCPP polymer. (A) Images have been recorded before skin application and (B) 4 h after. The hair appears in green due to the autofluorescence and the collagen fibers in grey which corresponds to the second harmonic generation (SGH). (C) Montage of the Z projection from the bottom to the top.

b. Successful delivery of the vaccine antigen to skin dendritic cells

In a first attempt, we investigated whether these polymers are able to protect the cargo and deliver it to the skin DC. To answer this question, we performed an adoptive transfer experiment in which OVA protein has been formulated with either PCPP or PCEP and subsequently applied on the skin of mice preliminary injected with CFSE-labeled OVA-specific T-lymphocytes. Since it has been demonstrated by Mittal et al. the need to include adjuvants to generate efficient antigen-specific humoral and cellular immune responses when using purified antigens and NP for TF vaccination119, we performed this adoptive

59

transfer experiment using a co-formulation of antigen/polymer with adjuvant. In figure 23it is shown that both polymers are able to deliver OVA, thereby promoting the stimulation of an OVA-specific adaptive immune response. However, the PCEP formulation seems to be more efficient, since significantly stronger CD4 and CD8 immune responses are stimulated with respect to those observed in the control group of non-immunized animals. Interestingly, while PCEP stimulated predominantly CD4+ T-cells at both local and systemic level, the systemic CD4 response seems to be diminished when using PCPP (Fig. 23B).

Figure 23: Proliferation of OVA-specific CD4 and CD8 T-cell lymphocytes. OVA-specific naive T-cells were labeled and adoptively transferred into WT mice 24 h prior immunization. Data represent the percentage of proliferating labeled cells in the draining LNs (A) and in the spleen (B). Differences are statistically significant with P<0.05 (*) and P<0.01 (**) compared to the control group).

c. Transfollicular immunization using polyphosphazenes as nanocarrier stimulates antigen-specific immune responses

Having validated that polyphosphazene polymers are able to protect and deliver the vaccine antigen into the HF, we then investigated if the formulations are able to evoke efficient immune responses against the antigen in naïve mice which are able to protect them against a virus challenge. It has been previously demonstrated that using chitosan/PLGA NP by TF route resulted in the stimulation of strong antigen-specific immune responses and that adjuvantion is required to stimulate optimal responses119. Thus, we characterized the

60

adaptive immune response stimulated by a c-di-AMP adjuvanted formulation of OVA-polyphosphazenes in comparison to a formulation based on chitosan/PLGA (CS/PLGA) NP, which was used as golden standard. When evaluating the systemic humoral immune response stimulated after four immunizations we observed that polyphosphazene NP are able to induce ten-fold higher IgG titers as compared to the titers obtained using chitosan/PLGA NP (Fig. 24B). No statistically significant differences could be observed between PCEP and PCPP. Interestingly, unlike what has been previously published, here the polyphosphazenes seem to be more efficient at stimulating the production of the IgG2c subclass with respect to the CS/PLGA NP (Fig. 24C).

OVA

-spe

cific

IgG

tite

r in

sera

(e.d

.p)

Control

OVA-PCPP + c-di-A

MP

OVA-PCEP + c-di-A

MP

OVA-CS/PLGA + c-d

i-AMP

100

1000

10000

100000

1000000 *

OVA-PCPP + c-di-A

MP

OVA-PCEP + c-di-A

MP

OVA-CS/PLGA + c-d

i-AMP

100

1000

10000

100000

1000000 IgG1IgG2c

B C

Figure 24: Systemic humoral immune responses induced in mice after four immunizations with OVA-polyphosphazene polymers via TF route. (A) Vaccination schedule (200 µg OVA per dose). (B) OVA-specific IgG titers in sera 13 days after the last immunization. (C) OVA-specific IgG1 and IgG2c in the sera of immunized mice.

Prime 1st Boost 2nd Boost Samling

-2 14 28 560 12 26

Shave Shave Shave

Vaccination schedule

3rd Boost

4240

Shave

A

61

Results are displayed as average of the last dilution of sera (endpoint dilution, e.p.d) showing the double value (OD405nm) of the control background. SEM is indicated by vertical lines, (n=12). Differences are statistically significant with P<0.05 (*) compared to the control group.

In line with the results obtained by Mittal et al.119 CS/PLGA NP in combination with c-di-AMP induce a strong antigen-specific cellular response, as indicated by the proliferative capacity of the splenocytes upon restimulation with increased concentrations of OVA, reaching an index of 4 (Fig. 25A). In contrast, similar levels have been obtained using PCEP only when splenocytes were restimulated with high concentrations of OVA, whereas the PCPP formulation induced only a marginal proliferative capacity, if any at all. When analyzing in more in detail the cell types proliferating, it seems that the CS/PLGA NP are more efficient in stimulating CD8+ T-cells as compared to the polyphosphazene formulations (Fig. 25B).

Control

OVA-PCPP + c-di-A

MP

OVA-PCEP + c-di-A

MP

OVA-CS/PLGA +

c-di-A

MP

Figure 25: Cellular immune response stimulated by c-di-AMP adjuvanted OVA-polyphosphazene formulations. (A) Splenocytes from vaccinated animals were restimulated with increasing concentrations of the OVA protein, the proliferation of the cells after 96 h was then determined by the incorporation of 3H-thymidine. Results are the average of quadruplicates and are expressed as SI. Statistically significant differences compared to the control group were validated using two-way ANOVA with Dunnett’s multiple comparison test, with P < 0.05 (*), P < 0.01 (**), P < 0.001 (***) and P < 0.0001 (****). (B) Cells derived from draining LNs from vaccinated animals were stained with the CFSE dye and restimulated with 10 µg/ml of OVA protein for 96 h. The proliferation capacity of the cells was assessed by FACS. Data are expressed as number of antigen-specific proliferating cells per 500000 LNs cells. Draining LNs cells from mice belonging to the same group (n=3) were pooled.

62

Nevertheless, the PCEP and PCPP formulations are able to stimulate a significantly increased number of IFNγ producing CD4+ or CD8+ T-cells as compared to CS/PLGA, whereby PCEP seem to be more efficient at stimulating IFNγ producing OVA-specific CD8+ T-cells (Fig. 26A). On the other hand, CS/PLGA stimulated higher numbers of antigen-specific IL-2 and IL-17 producing cells compared to both PCEP and PCPP (Fig. 26A). Considering the frequency of multifunctional CD4+ T-lymphocytes, no statistical differences have been observed between the three formulations. However, when analyzing each bi-functional (INFγ/IL-2, IFNγ/TNFα and IL-2/TNFα) and tri-functional (IFNγ/TNFα/IL-2) CD4+ T-cells, PCPP and PCEP show a similar pattern, whereas CS/PLGA induced a higher frequency of IL-2/TNFα and less IFNγ/TNFα bi-functional T-cells as compared to the two polyphosphazene formulations (Fig. 26B). All in all, using polyphosphazenes as a delivery system for protein antigens seems to induce a more balanced Th1/Th2 immune responses as compared to CS/PLGA NP, but does not increase the frequency of multifunctional CD4+ T-lymphocytes.

63

Figure 26: (A) The number of cytokine-producing cells (IFNγ, IL-2, IL-4, and IL17) was determined by ELISpot. Splenocytes were restimulated with 10 µg/ml of OVA protein (2 µg/ml of the SIINFEKL peptide). Results are expressed as spot-forming units of 1x106 restimulated cells. The background was removed using unstimulated samples. SEM is indicated by vertical lines, (n=3). (B) The histogram displays the frequency of multifunctional CD4+

T-lymphocytes in the spleen 24 h after in vitro antigen stimulation and determined by flow cytometry analysis. The pie charts represent the percentage of each bi- (IFNγ/IL-2, IFNγ/TNFα, IL-2/TNFα) or tri-functional (IFNγ/IL-2/TNFα) CD4+ T-cells among the total multifunctional cells. Background was removed using unrestimulated samples. SEM is indicated by vertical lines, (n=3). For each group differences are statistically significant with P<0.05 (*), P<0.01 (**), P<0.001 (***) and P<0.0001 (****) compared to the group OVA-CS/PLGA + c-di-AMP.

In order to establish if vaccinated mice would be able to detect virus-infected cells and lyse them, we performed an in vivo CTL study. It showed that, although very modest, the PCPP formulation seems to be the most efficient approach for stimulating CTL compared to the other formulations (Fig. 27).

64

Control

OVA-PCPP + c-di-A

MP

OVA-PCEP + c-di-A

MP

OVA-CS/PLGA +

c-di-A

MP

Positive

contro

l0

2

4

6

8

1040

50

6070

80

SpleenLymph nodesLungs

************

****

Figure 27: In vivo cytotoxic lymphocytes killing assay. The function of CTLs stimulated after TF vaccination has been monitored in an in vivo assay. Results are expressed as a percentage of specific killing in the spleen, LNs and lungs.

d. Transfollicular immunization using formulations based on polyphosphazenes protects mice against live virus challenge

Having demonstrated that the immune cells of all groups of vaccinated mice are able to specifically react when restimulated ex vivo with the corresponding antigen, we next investigated the ability of the different vaccine formulations to confer protection of the mice against infection with a live influenza virus expressing the OVA CD8 peptide SIINFEKL. Unfortunately, all vaccinated animals showed a loss of weight similar to the non-vaccinated mice, starting at day 3 after the challenge, peaking at day 7 and reaching a normal weight at day 10. In contrast, mice of the positive control (vaccinated with adjuvanted OVA protein by i.n. route) showed no weight loss (Fig. 28). These results confirm those of the in vivo CTL tests (i.e. poor capacity of the polyphosphazenes to stimulate CTLs), since protection in this particular challenge model is only conferred by CTLs recognizing the SIINFEKL peptide.

65

Perc

ent o

f Wei

ght

Figure 28: Protection of vaccinated animals against influenza A virus infection. (A) C57BL/6 mice (n=6) received four immunizations with adjuvanted OVA-polyphosphazene formulation by TF route. Three weeks after the last immunization, mice were challenged with a sub-lethal dose (3x102 ffu) of the influenza strain A/PR8/OVA SIINFEKL CD8 (H1N1) given by i.n. route (10 µl per nostril). The animal body weight was then monitored for a period of 14 days. Results are expressed as the average percentage of body loss.

4. Polyphosphazene polymer-based formulation for the recombinant HA protein

a. Recombinant HA protein efficiently activate antigen-specific CD4+ and CD8+

T-cells when applied via the TF route

As mentioned above, although the immune responses stimulated after TF vaccination using OVA polymer seemed to be qualitatively and quantitatively strong enough to expect some degree of protection against influenza virus challenge, no immunity have been observed. However, the challenge model is based on an influenza virus expressing only the OVA

66

CD8 peptide SIINFEKL. Therefore, only one arm of the immune system, the CD8+ T-lymphocytes, could have acted for the clearance of the virus, whereas OVA-specific antibodies and T helper cells were useless. Thus, we decided to use recombinant hemagglutinin protein (HA) from influenza for further studies. In addition, we included αGCMPEG as adjuvant as we have also demonstrated its potential for TF immunization when using inactivated H1N1 as vaccine antigen (see Results section 2). As for the formulations encompassing OVA protein, we first performed a proof-of-concept study in order to validate the usefulness of the formulation containing recombinant HA. To this end, mice that have preliminarily received an i.v. injection of CFSE labeled HA-specific CD4+ and CD8+ T-lymphocytes were immunized by TF route using 25 µg of HA formulated with polyphosphazenes with and without co-administration of the adjuvants c-di-AMP and αGCMPEG, respectively. Seven days after TF application of the different vaccine formulations, we analyzed the proliferation rate of the injected cells. In line with the results obtained using OVA as antigen, TF immunization with the recombinant HA protein formulated with polyphosphazenes also results in the activation of HA-specific CD4+ and CD8+ T-cells. Thus, strong proliferation rates of CD4+ and CD8+T-cells have been observed both in the draining LNs and the spleen of immunized mice (Fig. 29A and B). For all formulations used the observed proliferation rates were significantly stronger compared to those of cells derived from mice of the control group. Interestingly, at systemic level co-administration of HA-PCEP with c-di-AMP seems to be more efficient at stimulating the proliferation of CD4+ T-cells with respect to all the other formulations. In contrast, HA-PCPP co-administered with αGCMPEG seems to be most efficient in stimulating CD8+ T-cells (Fig. 29B). However, the differences observed between the two polyphosphazene based formulations were not statistically significant. The presence of the adjuvants seemed to be dispensable, since similar results were obtained using adjuvanted and non-adjuvanted formulations. However, it is important to highlight that this needs to be interpreted cautiously, since in this model the number of antigen-specific T-cells is approximately 200-times higher than in naïve animals (i.e. a considerable lower efficacy is expected under a true vaccination setting).

67

Freq

. of p

rolif

erat

ing

cells

Control

HA-PCPP

HA-PCEP

HA-PCPP + c-di-A

MP

HA-PCEP + c-di-A

MP

HA-PCPP + GCMPEG

HA-PCEP +GCMPEG

Control

HA-PCPP

HA-PCEP

HA-PCPP + c-d

i-AMP

HA-PCEP +c-d

i-AMP

HA-PCPP + GCMPEG

HA-PCEP +GCMPEG

Figure 29: HA-specific proliferation of CD4 and CD8 lymphocytes following TF vaccination using recombinant HA protein co-formulated with the polyphosphazene polymers PCPP or PCEP. HA-specific naive T-cells were adoptively transfered into WT mice before a single immunization with HA-polymer with or without c-di-AMP or αGCMPEG. Data represent the percentage of proliferating cells in (A) the draining LNs and (B) the spleen. Differences are statistically significant with P<0.05 (*), P<0.01 (**), P<0.001 (***) and P<0.0001 (****) compared to the control group.

b. Characterization of the adaptive immune responses

Like in the case of OVA, the HA protein is also successfully delivered to the skin APC using polyphosphazenes as delivery system, resulting in the activation of HA-specific T-cells. This in turn also indicates a preserved structure of the recombinant HA protein, and adequate deliver to APC and subsequent uptake, processing and presentation to HA-specific T-cells. Therefore, we performed a side-by-side comparison of the two polymers PCPP and PCEP alone and co-administered with c-di-AMP or αGCMPEG to characterize in detail the HA-specific immune responses stimulated in naïve mice. Considering the results obtained from the proof-of-concept studies (i.e. an increased immunogenicity of HA compared to OVA, see section 5a) as well as the production costs of the recombinant HA protein, the antigen dose was reduced by a factor 4 as compared to the experiments using OVA as vaccine antigen (i.e. 50 µg of HA protein per animal and per dose). Thus, we performed a vaccination experiment following the same protocol described in section 4a iii, including one prime followed by three booster immunizations. Two weeks after the

68

last boost, the IgG antibody titer revealed that c-di-AMP and αGCMPEG significantly enhanced the HA-specific antibody titer when co-formulated with PCEP polymers as compared to the non-adjuvanted PCEP. In contrast, for PCPP no significantly increased antibody titer has been obtained when formulated with either of the two adjuvants (Fig. 30A). Investigating the antibody subclasses stimulated revealed that HA-polyphosphazenes co-formulated with c-di-AMP induced a more balanced IgG1/IgG2a antibody ratio (100) as compared to the non-adjuvanted formulations (1000), indicating a balanced Th1/Th2 immune response. On the contrary, co-administration of αGCMPEG does not alter the IgG subclass ratios stimulated by the polymers alone with only very low titers of IgG2a, indicating a Th2 dominated immune response (Fig. 30B).

HA-PCPP

HA-PCEP

HA-PCPP + c-d

i-AMP

HA-PCEP + c-di-A

MP

HA-PCPP +GalC

erMPEG

HA-PCEP +GalC

erMPEG

Figure 30: Systemic humoral immune responses induced in mice after four immunizations with polyphosphazene polymers co-formulated with the recombinant HA protein via TF route. (A) HA-specific IgG titers in sera 13 days after the last immunization. (B) HA-specific IgG1 and IgG2a titers in sera 13 days after the last immunization. Results are displayed as the average of the last sera dilution (endpoint dilution, e.p.d) showing the double value (OD405nm) of the control background. SEM is indicated by vertical lines, (n=6 for the group control, HA-PCPP + c-di-AMP; n=9 for the group HA-PCPP, HA-PCEP, HA-PCEP + c-di-AMP, HA-PCEP + αGCMPEG and n=3 for the group HA-PCPP + αGCMPEG). Differences are statistically significant with P<0.05 (*), P<0.01 (**) and P<0.001 (***) compared to the corresponding non-adjuvanted group.

69

As proven for the humoral responses obtained in mice vaccinated with the polyphosphazene-based formulations, also the strength of the cellular immune responses strongly depends on the used polymers and adjuvants. Thus, HA-PCPP co-administered with or without adjuvant stimulated strong cellular responses as indicated by increased proliferation rates after re-stimulation with not more than 1 µg/ml of the corresponding antigen (Fig. 31). However, when restimulating with higher antigen concentrations becomes obvious that incorporation of c-di-AMP or αGCMPEG further increased the proliferative capacity of splenocytes derived from mice vaccinated with the HA-PCPP formulation (Fig. 31). Thus, at the maximal antigen concentration (20 µg/ml) all four adjuvanted formulations reach a stimulation index between 4 and 5, whereas the HA-polymers without adjuvant peak at 3 (Fig. 31). Similar results have been observed in case of the HA-PCEP formulation. However, the obtained proliferation rates are significantly lower as compared to non-adjuvanted HA-PCEP (Fig. 31). Moreover, no significant differences have been observed between the adjuvanted formulations.

Figure 31: Cellular immune response stimulated by adjuvanted HA-polyphosphazene formulations. Splenocytes from vaccinated animals were restimulated with an increased concentration of HA protein (starting from 1 µg to 20 µg/ml). Proliferation of the cells after 96 h was determined by the incorporation of 3H-thymidine. Results are average of quadruplicates and are expressed as SI. SEM is indicated by vertical lines. Statistically significant differences compared to the corresponding non-adjuvanted group with respect to HA-PCPP, HA-PCEP, HA-PCPP + c-di-AMP (+), HA-PCEP + c-di-AMP (I), HA-PCPP + αGCMPEG (o) or HA-PCEP + αGCMPEG (□) were evaluated using two-way ANOVA with Dunnett’s multiple comparison test, with P< 0.05 (*), P< 0.01 (**), P< 0.001 (***) and P< 0.0001 (****).

70

Upon in vitro antigen restimulation, lymphocytes of vaccinated animals do not only proliferate, they also secrete cytokines which allow to draw conclusions about their polarization. In line with the observed IgG subclass profiles, HA-polyphosphazene formulations seem to stimulate a predominant Th2 like immune response, as demonstrated by increased numbers of IL-4 producing cells (Fig. 32A). The incorporation of c-di-AMP and αGCMPEG in the formulation resulted in significantly increased numbers of HA-specific IL-2 and IL-4-producing cells only when using the PCEP polymer, as compared to HA-PCEP alone. In addition, the number of IFN-producing cells is significantly higher when mice were vaccinated with HA-PCEP co-administered with c-di-AMP and αGCMPEG, as compared to those detected in mice receiving HA-PCPP + αGCMPEG and c-di-AMP (Fig. 32A). However, when analyzing the cytokine concentrations detected in the supernatant of antigen restimulated splenocytes from immunized mice, polyphosphazenes also seem to stimulate Th17 cells as indicated by increased levels of IL-17A and IL-22 (Fig. 32B). The same is true when HA-polymers were co-administered with αGCMPEG. Again, consistent with the observed IgG subclass profiles, incorporation of αGCMPEG did not alter the Th cell polarization nor increased cytokine production (Fig. 32). Interestingly, αGCMPEG is known to suppress the stimulation of Th17 responses favoring more a Th1 response136,137, but in case of the HA-polymers it could not overcome the responses stimulated by the polyphosphazene. The c-di-AMP seems to further increase the cytokine production stimulated by HA-PCPP, whereas almost no cumulative effect could be observed for the HA-PCEP formulation (Fig. 32B). Notably, PCPP seems to increase the amount of IFN produced by splenocytes derived from immunized mice, whereas TF application of PCEP results in increased numbers of IFN producing cells (Fig. 32).

71

Figure 32: Antigen-specific cytokine-profiles stimulated by HA-polymers co-administered with different adjuvants. (A) The number of cytokine-producing cells (IFNγ, IL-2, IL-4, and IL17) was determined by ELISpot. Splenocytes were restimulated with 10 µg/ml of HA protein. Results are expressed as spot-forming units of 1x106 restimulated cells. Background was removed using unstimulated samples. SEM is indicated by vertical lines, (n=3). Differences are statistically significant with P<0.05 (*), P>0.01 (**) and P>0.001 (***) as compared to the corresponding non-adjuvanted group. (B) Heatmap representing the cytokines secreted by splenocytes (pg/ml) derived from vaccinated mice incubated with recombinant HA [10 µg/ml] for 96 h.

A

B

72

c. Transfollicular immunization using polyphosphazene polymer formulated with recombinant HA protein protects mice against influenza infection

Characterization of the adaptive immune responses stimulated following HA-polyphosphazene vaccination revealed small quantitative and qualitative differences between the two polymers PCPP and PCEP. This is especially true for the stimulated cellular responses. Therefore, in the next step we evaluate if the HA-polyphosphazene formulations applied via TF route are able to confer protection against an influenza virus infection. Currently, it is well accepted that neutralizing antibodies correlate (imperfectly) with protection138–141. However, promoting robust immune defenses, including cytotoxic immunity, would provide more efficient protection against influenza virus, for both defending against virus antigenic change (drift and shift) and longevity of protective immunity. Thus, pre-existing influenza-specific CD4+ T-cells correlate with disease protection against experimental influenza challenge in man, whereas virus-specific cytotoxicity inversely correlated with the extent of virus shedding after experimental infection in the absence of influenza-specific antibodies142,143. Therefore, we next investigated initially the protective capacity of the immune responses stimulated by HA-PCEP co-administered with c-di-AMP and αGCMPEG. Following TF immunization, mice were infected three weeks after the last immunization with the influenza strain A/PR8 H1N1 expressing the luciferase. Mice vaccinated with HA-PCEP co-administered with either of the adjuvants were fully protected (100%) against the influenza infection, not showing any malaise as indicated by their constant weight levels (Fig. 33A and B). In contrast, TF immunization of mice using HA-PCEP alone did not stimulate efficient immune responses resulting in only 33.3% protection compared to 16.6% protection observed for the untreated control group (Fig. 33B).

73

Figure 33: Protection of vaccinated animals against influenza A virus infection. (A) BALB/c mice (n=6) received four immunizations with adjuvanted or non-adjuvanted HA-PCEP formulation by TF route (the positive control was vaccinated s.c. with 10 µg of c-di-AMP adjuvanted HA protein). Three weeks after the last immunization, mice were challenged with the influenza strain A/PR8/Nluc (H1N1) given by i.n. route (10 µl per nostril). Animal health status was monitored for a period of 14 days. (A) Weight development of infected mice is expressed as the average percentage of body loss. (B) Survival rate is expressed as percentage of survival.

Experiments were also performed to evaluate the viral loads in lungs of immunized mice following infection with the influenza strain A/PR8/Nluc (H1N1). This influenza virus contains the sequence for the luciferase Nluc downstream of the PA polymerase subunit sequence allowing in vivo monitoring of the course of infection144. In line with the observed survival rates, non-vaccinated mice (control group) and mice vaccinated with HA-PCEP alone showed higher viral loads in the lungs compared to mice receiving HA-PCEP co-administered with an adjuvant, as demonstrated by the luminescence intensity recorded in the lungs at day 7 after i.n. infection (Fig. 34A and B). Nevertheless, although TF immunization of mice using HA-PCEP co-administered with c-di-AMP and αGCMPEG results in 100% protection against influenza it seems that the efficacy of this vaccination strategy could still be optimized. Hence, estimating that 10% of the applied antigen dose enters the skin already a twofold increase of the dose seems to be efficient in further reducing viral load as indicated by the reduced viral load in lungs of mice immunized by s.c. route using 10 µg of HA (Fig. 34A and B).

74

Figure 34: In vivo bioluminescence imaging of mice at day 7 post-inoculation of live H1N1 virus expressing the luciferase protein. (A) A representative image series of four vaccinated mice from each group. Data are expressed in photons per second per square centimeter per steradian (p/s/cm2/sr). (B) Graph summarizing the average of radiance recorded by the in vivo imaging system. SEM is indicated by vertical lines (n=4). Differences are statistically significant with P>0.01 (**) compared to the control group.

5. Deciphering the local early events triggered by transfollicular immunization.

a. Antigen uptake by skin DC following TF immunization

Although we have demonstrated that TF vaccination results in the induction of potent antigen-specific immune responses protecting mice against an influenza infection, we aimed at unraveling the underlying mechanisms in order to further improve vaccine efficacy. To this end, we initially identify the skin DC subset targeted by TF vaccination. For this, naïve C57BL/6 mice were immunized on the ventral ear side via the TF route using DQ-OVA, a fluorescent-labeled OVA protein in which the fluorescence is emitted upon cellular processing of the protein antigen. As it has been already described the antigen

75

needs to be protected against skin protease degradation, but also a certain size (around 180 µm) of the antigen is needed to reach a maximal depth of the HF through mechanical interactions with the hair50,51. For this reason, we decided to use the already known polyphosphazene polymer family and more precisely the PCPP member, which will form with OVA a nanoparticle like formulation. In addition, considering the quantitative and qualitative improvement of the immune responses when the inactivated H1N1 was co-formulated with either c-di-AMP or αGCMPEG, we decided to include both of them in this study. Therefore, four formulations have been compared side by side: DQ-OVA alone, DQ-OVA formulated with PCPP, DQ-OVA PCPP adjuvanted with c-di-AMP and DQ-OVA PCPP adjuvanted with αGCMPEG. Ears were harvested at different time points between 1 to 96 hours after immunization, skin cells and draining LNs were subsequently isolated and processed for flow cytometric analysis (Fig. 35).

As shown in figure 36A, antigen uptake and processing by skin DC is strongest already 1 h post-immunization. For all the formulations, the percentage of DQ-OVA positive DC is similar (6 to 7%). However, DC targeted by DQ-OVA PCPP co-administered with c-di-AMP seem to migrate earlier to the draining LNs as compared to DQ-OVA PCPP alone or co-administered with αGCMPEG adjuvant (Fig. 36A). Thus, while in mice receiving c-di-AMP as adjuvant almost no DQ-OVA positive DC could be detected in the skin already after 24 h, this was the case for mice receiving DQ-OVA PCPP alone or adjuvanted with αGCMPEG only 72 h after TF application. By calculating the area under the curve (AUC) of each vaccine formulation, which corresponds to the sum over time of the cumulated cell abundances expressing the antigen from 0 h to 96 h, we can confirm that the DC targeted

Immunization Sampling

3 24 96 hours0 1 156 7248

SamplingSamplingSamplingSampling

Sampling

Sampling

Sampling

Figure 35: Immunization and sampling schedule.

76

by DQ-OVA PCPP + αGCMPEG adjuvant formulations persist significantly longer in the skin before migrating to the draining LNs as compared to the DC from mice immunized with DQ-OVA alone (Fig. 36B).

DQ-OVA

DQ-OVA PCPP

DQ-OVA PCPP + c-d

i-AMP

DQ-OVA PCPP +

GCMPEG

Figure 36: Kinetic of vaccine antigen uptake and processing by skin DC. (A) Frequence of antigen-positive skin DC at the site of vaccine application was determined by flow cytometry over time (n=6 for each time point from 0 to 48 h for groups DQ-OVA PCPP, DQ-OVA PCPP + c-di-AMP and DQ-OVA PCPP + αGCMPEG and n=3 for group DQ-OVA; n=3 for time points 72 h and 96 h for the group DQ-OVA PCPP, DQ-OVA PCPP + c-di-AMP, and DQ-OVA PCPP + αGCMPEG); no data for the time point 72 h and 96 h for the group DQ-OVA. (B) The AUC of each formulation is displayed (AUC of the frequency of antigen-positive DC of the whole kinetic with a baseline of 0). SEM is indicated by vertical lines. Differences are statistically significant with P<0.05 with respect to the group DQ-OVA alone.

We then determined more in detail which skin DC subpopulation is prone to uptake DQ-OVA antigen delivered using PCPP alone or adjuvanted. As shown in figure 37, the majority of DQ-OVA is taken up by the CD11b+ DCs population (between 86 to 96%), independently of the used formulation. Another 7% to 10% are taken up by the LC subpopulation, whereby co-administration of αGCMPEG seems to enhance the frequency of uptake by LC as compared to DQ-OVA delivered using PCPP alone (Fig. 37). In

77

contrast, DN-DCs and XCR1+ DCs represent only 1 to 1.5% of the skin DCs taking up DQ-OVA (Fig. 37). In line with what was observed for the overall kinetic, this holds true over 6 h prior to cell migration to the draining LNs (Fig. 37A-C).

+ +

XCR1+ DCs

LCs

CD11b+ DCs

DN cDCs

+ +

Figure 37: Vaccine antigen uptake and processing by skin DC subpopulations. The frequency of each of the four skin DC subpopulations among all the DQ-OVA positive DCs is displayed for all PCPP formulations. (A) 1 h, (B) 3 h and (C) 6 h post-immunization. SEM is indicated by vertical lines (n=6). Differences are statistically significant with P<0.05 (*) and P>0.01 (**).

b. DC activation following transfollicular immunization

In the next attempt, we analyzed the activation status of the skin DC subpopulations following TF immunization using DQ-OVA loaded PCPP polymer with or without co-administration of an adjuvant. The obtained results show that all formulations stimulated the same degree of expression of the activation marker CD86 by LC even when mice received DQ-OVA alone (Fig. 32). In contrast, c-di-AMP seems to be most efficient in activating CD11b+ DCs, DN cDCs, and XCR1+ DCs, as indicated by increased CD86 expression levels as compared to all other vaccine formulations. However, the differences were significant only in the case of the CD11b+ DCs and DN cDCs populations when compared to those obtained in mice receiving DQ-OVA alone (Fig. 38). Interestingly, co-administration of αGCMPEG does not seem to further increase the levels of expression of CD86 with respect to what was observed for DQ-OVA PCPP alone (Fig. 38).

78

CD8

6 ex

pres

sion

nor

mal

ized

to 0

h (A

UC)

Figure 38: Activation of skin DC after TF immunization of mice with different PCPP based formulations. The expression level of the activation marker CD86 was evaluated by flow cytometric analysis (MFI) and results were displayed as AUC of cumulated abundance of CD86 expression over time from 0 h to 96 h. SEM is indicated by vertical lines. Differences are statistically significant with P<0.05 (*) compared to values obtained for mice receiving DQ-OVA alone.

c. Migration of skin dendritic cells.

We next evaluated the migration properties of the skin DC following TF vaccination using DQ-OVA–PCPP formulations. To this end, the frequency of the four DC subpopulations of the skin was monitored by flow cytometry for a period of 96 h. The acquired values were normalized for those detected at timepoint 0 h. The obtained results indicate, that CD11b+ DCs are recruited to the area of application starting 6 h after immunization, independently of the used formulation (Fig. 39A). In contrast, DN cDCs start to migrate and leave the skin 15 h post-immunization as indicated by the decreasing frequency (Fig.

79

39B). Similar migration behaviors have been observed in case of the XCR1+ DCs. However, this subpopulation seems to migrate already after 1 h, especially when mice received the formulation including c-di-AMP as adjuvant (Fig. 39D). Interestingly, LC do not seem to migrate but rather persist in the skin (Fig. 39C).

Figure 39: Migration kinetics of the four main skin DC subpopulations. Frequency of each subpopulation, for each time point and for each formulation, determined by flow cytometry among CD11c+ MHC-II+ cells (the frequency is normalized to the timepoint 0 h among each experiment). Data are average of 6 mice out of two experiments for the time points 0 h to 48 h regarding the formulations containing PCPP and 3 mice for the formulation DQ-OVA alone. For the time points 72 h and 96 h three mice for all the formulation with PCPP have been used while no data for the group DQ-OVA alone have been investigated.

In order to investigate more in detail the impact of each vaccine formulation on the four DC subpopulations, the area under (recruitment) and above (migration) the curve, respectively, has been calculated for each formulation considering a baseline of 1 (Fig. 40).

80

Again, the obtained results showed no differences between the migration behavior of CD11b+ DCs stimulated with the different formulations. Thus, all PCPP based formulations significantly increased the recruitment of this DC subpopulation as compared to the formulation encompassing DQ-OVA alone for which no migration could be observed. When evaluating the impact of the PCPP on LCs, recruitment and migration are balanced but slightly increased compared to DQ-OVA alone (Fig. 40). In contrast, although not statistically significant, using PCPP with or without co-administration of an adjuvant a clear trend emerged suggesting enhanced migration capacity of DN cDCs, especially in case of αGCMPEG. Interestingly, the migration capacity of XCR1+ DCs was significantly stronger in mice receiving DQ-OVA PCPP formulation co-administered with c-di-AMP, whereas αGCMPEG had almost no effect on this DC subpopulation as compared to DQ-OVA alone (Fig. 40).

Figure 40: Impact of the formulation on the frequency of skin DC subpopulations. The frequency of each DC subpopulation was monitored over a period of 96 h and expressed as AUC. For frequencies increasing over time, AUC was considered as recruitment into the skin (white bars) whereas for frequencies decreasing over time AUC was considered as migration out of the skin (black bars). SEM is indicated by vertical lines (n=6 for all formulations except

81

for DQ-OVA group (n=3)). Differences are statistically significant with P<0.05 (*) and P<0.01 (**) compared to DQ-OVA alone group.

d. Migration of skin dendritic cells to the draining lymph nodes

After analyzing the migration behavior of the skin DC subpopulations following TF immunization with the PCPP based vaccine formulations, we investigated the draining LNs of the vaccinated mice for influx of migratory DC. Migrating skin DC easily can be distinguished from LN-resident DC because of their increased expression level of the MHC-II molecule on their surface145. Analyzing by flow cytometry the frequency of migratory DCs in the draining LNs during our kinetic schedule, we observed that TF immunization with DQ-OVA PCPP + c-di-AMP resulted in significantly increased numbers of migratory DC in the draining LNs when compared to DQ-OVA alone (Fig. 41A). Although the obtained levels of migratory DC using c-di-AMP were also increased compared to DQ-OVA PCPP alone and co-administered with αGCMPEG, respectively, the differences were not significant (Fig. 41A). The same is true when comparing the expression levels of the well-known DC activation marker CD86. However, 48 h after immunization, upregulation of CD86 by the cells migrating from the skin to the draining LNs is significantly higher in mice receiving DQ-OVA PCPP adjuvanted with c-di-AMP as compared to those of mice immunized with DQ-OVA alone and DQ-OVA PCPP (Fig. 41B).

Figure 41: Recruitment of migratory skin DC following TF immunization. (A) The frequency of the skin migratory DCs is displayed as AUC for each formulation. (B) The expression of CD86 activation marker, normalized to the time point 0 h is displayed as curve over time. SEM is indicated by vertical lines (n=6 for all formulations except for DQ-

82

OVA group (n=3)). Differences are statistically significant with P<0.05 (*), P<0.01 (**) and P<0.0001 (****) compared to DQ-OVA alone group.

e. Deeper characterization of the migrating skin dendritic cells

Although migratory skin DC can be identified within the draining LNs, not all of these DC also carry the antigen. In addition, cells are not migrating at the same time and some migrating cells also tend to change their phenotype upon migration, such as the XCR1+

DCs for example35,146. Therefore, instead of verifying the different migratory DC subsets in the draining LNs in vivo we decided to deeper characterize the migration behavior of skin DC subsets post-immunization using an ex vivo model of skin explant cultures. In that case, LC and XCR1+ DCs cannot be distinguished with the expression of CD11 marker, therefore in the next experiment we have analyzed both populations together based on the marker CD24 only and referred them as langerin positive cells. According to the previous data, we have decided to study the migration of the skin cell ex vivo for 72 h after TF immunization.

Surprisingly, when focusing on the antigen-positive migratory cells, Langerin+ cells turned out to be the most common cells migrating to the draining LNs upon vaccine antigen uptake followed by CD11b+ cDC and DN-DC (Fig. 42). Furthermore, co-administration of c-di-AMP significantly increased the number of antigen-positive migratory Langerin+ cells as compared to non-adjuvanted PCPP formulated DQ-OVA. However, no statistically significant differences could be observed between the adjuvanted and non-adjuvanted groups for the CD11b+ cDC and DN-DC subpopulations (Fig. 42).

83

CD11b+ cD

Cs

Langeri

n+ ce

lls

DN-DCs

Figure 42: Migration of antigen-positive DC subpopulations. Data are the number of antigen-positive cells that migrate from ear skin over a period of 72 h of culture. SEM is indicated by vertical lines (n=6 for all formulations except for DQ-OVA group (n=3)). Differences are statistically significant with P<0.05 compared to non-adjuvanted DQ-OVA PCPP group.

f. Local pro-inflammatory chemokine profile

So far we know that TF immunization with the different PCPP based vaccine formulations results in migration and recruitment of DC. In order to get a more detailed understanding of the underlying mechanisms we also investigated the local microenvironment stimulated upon TF immunization. To this end, we analyzed the kinetics of the pro-inflammatory chemokine profiles in the skin tissue stimulated following TF immunization. In figure 43is displayed the fold change increase of the chemokine concentrations in the skin over time for three DQ-OVA PCPP based formulations. Interestingly, depending on the composition of the used formulation very different chemokine profiles have been observed (Fig. 43). Indeed, using the non-adjuvanted DQ-OVA PCPP formulation only a few chemokines were upregulated starting 6 h post-immunization. Thus, the expression of the chemokines CXCL-1 and CCL-2 (MCP-1), which are responsible for chemotactic and proinflammatory effects, was raised at 6 h and lasted for 18 h. On the other hand, the production of CCL-3 (MIP-1α), which specifically attracts immature DC and lymphocytes, was strongest 72 h after immunization (Fig. 43A). In contrast, when c-di-AMP was incorporated not only much higher chemokine expression levels but also a broader spectrum of chemokines have been observed post-immunization. In addition to CXCL-1, CCL-2 and CCL-3, also expression of the chemokines CCL-4 (MIP-1β, among others attracts natural killer cells), CCL-5 (among others chemotactic for T-cells) and CXCL-10 (chemoattractant for

84

different immune cells) have been strongly upregulated (Fig. 43B). Furthermore, expression of CXCL-1 and CCL-2 started already 3 h and 1 h after TF application, respectively. Nevertheless, for both formulations chemokine expression was strongest between 15 h and 24 h post-immunization (Fig. 43B). Surprisingly, mice receiving αGCMPEG as adjuvant mainly produced CCL-3 followed by CCL-2 and CXCL-1 (Fig. 43C). Furthermore, CCL-3 expression rapidly increased already 3 h after immunization and lasted for 4 days, indicating a prolonged specific attraction of immature DC and lymphocytes (Fig. 43C).

Figure 43: Pro-inflammatory chemokine profiles in the skin following TF vaccination monitored for 96 h. Data are displayed as fold change increase of the chemokine concentration [ng/ml] as compared to the time point 0 h.

g. Neutrophils recruitment in the skin

Recent studies have highlighted the role of neutrophils population during vaccination147,148. Indeed, neutrophils have been shown to be recruited very fast at the site of vaccine application. Moreover, Duffy et al. have observed the antigen uptake and migration behavior of the neutrophils when the modified vaccinia Ankara virus was injected intradermally149. Therefore, considering the release of proinflammatory chemokines in the skin, we then assessed the frequency of neutrophils recruited in the skin following immunization. In line with the observed chemokine profiles, TF immunization with adjuvanted and non-adjuvanted DQ-OVA PCPP induced the recruitment of neutrophils into the skin starting 6 h post-immunization and lasting up to 96 h, with a peak between 15 h to 24 h according to the formulation used (Fig. 44A). However, incorporation of c-di-AMP as adjuvant seems to be most efficient in the recruitment of neutrophils as compared

85

to the other formulations. Nevertheless, although not statistically significant, vaccination using DQ-OVA PCPP and DQ-OVA PCPP + αGCMPEG also seem to stimulate a stronger neutrophil recruitment as compared to the formulation containing only DQ-OVA (Fig. 44B).

0h 1h 3h 6h 15h

24h

48h

72h

96h

DQ-OVA

DQ-OVA PCPP

DQ-OVA PCPP + c-

di-AMP

DQ-OVA PCPP + GCMPEG

Figure 44: Kinetic of neutrophils recruitment in the skin following TF immunization. (A) Frequency of neutrophils among the CD45+ skin cells at the site of vaccine application was determined by flow cytometry over time (n=6 for each time point from 0 to 48 h for the groups DQ-OVA PCPP, DQ-OVA PCPP + c-di-AMP and DQ-OVA PCPP + αGCMPEG, and n=3 for group DQ-OVA; n=3 for time points 72 h and 96 h for the group DQ-OVA PCPP, DQ-OVA PCPP + c-di-AMP, and DQ-OVA PCPP + αGCMPEG); no data for the time point 72 h and 96 h for the group DQ-OVA. (B) The AUC of each formulation is displayed (AUC of the frequency of neutrophils population of the whole kinetic with a baseline of 1). SEM is indicated by vertical lines. Differences are statistically significant with P<0.05 (*) compared to values obtained for mice receiving DQ-OVA alone.

86

Discussion

Even before Jenner and Pasteur, physicians through every century used the most exotic inoculation techniques that finally led to the important development of vaccination strategies150. Nowadays, vaccines constitute one of the greatest success stories of medical interventions. For instance, vaccination resulted in the eradication of the smallpox virus and the control of ten major pathogens. Therefore, vaccination represents the most promising tool for the prevention and control of infectious diseases. However, despite this success, there is still plenty of room for improvement to face the ongoing and future epidemics and the spread of newly emerging pathogens. Furthermore, a number of infectious agents are still able to overcome vaccine efficacy and keep spreading151–153. For example, HIV viruses show a fast mutation rate and more strikingly infect those cells needed to efficiently combat this pathogen. Another example is the influenza viruses which are able to mutate and rearrange their epitopes in a short time usually making the vaccination effective for no more than a year.

On top of it, vaccination is currently facing another challenge, that comes from the population itself. Thus, there is a prevalent decline of compliance for vaccination, which is directly linked to the trust or mistrust of the population toward vaccine efficacy and safety. The resultant suboptimal vaccination rates subsequently lead to the rebound of some infections that were previously tightly under control154. For example, even though safe and cost-effective vaccines against measles are available, the gap in vaccination coverage leads to new outbreaks of the virus all over the world (Source WHO155). To prevent outbreaks, vaccination coverage of 95% is needed. However, public concern about measle vaccination explain that globally only 85% coverage is reached for the first dose of the measle vaccine, further dropping to only 67% for the second dose (Source WHO156). Taken together, there is an urgent need for the development of new safe and efficient vaccines with easy administration logistics to keep the success story of vaccination.

In this regard, we aim at developing a non-invasive vaccination strategy based on transdermal application. In the past few decades, cutaneous vaccine administration has been intensively studied. Besides being the largest organ of the human body, the skin tissue is populated by an enormous number of immune cells. Today, the skin is already used for drug delivery i.e. because of its interesting feature of the long-term reservoir of the HF157. Moreover, it has been shown that NPs are promising delivery systems for both transcutaneous vaccination and modulation of the stimulated immune responses according

87

to the delivery site158. Although most NPs are not able to cross the intact stratum corneum, once they are located in the HF, NPs and their cargo just need to overcome the stratum basal to get in contact with the skin immune cells159. Therefore, using skin tissue as a route for vaccine delivery is a promising alternative for current challenges in vaccine delivery. Furthermore, nanoparticle-based formulations have the advantage to be safe since biocompatible materials are used, usually easier to produce and often increase antigen uptake by immune cells, which can allow the reduction of the antigen dose leading to a cheaper vaccines77.

In the present work, we were able to provide the proof-of-concept for a non-invasive TF immunization strategy that stimulates protective immune responses against viral infection. TF vaccination of mice with inactivated influenza H1N1 virus (equivalent to 5 µg of HA protein) protected mice against an i.n. infection with the homologous influenza strain. Immunity was based on a strong antigen-specific antibody production and the stimulation of a mixed Th1/Th2 immune response. Similar protection using cholera toxin challenge have already been reported by Glenn et al. following transcutaneous immunization with the cholera toxin. Although this study showed an antibody titer at about 30 times higher than the one we observed using inactivated H1N1 virus, their immunization protocol appeared to be not fully non-invasive as the resting time between shaving and immunization was only 24 h. Indeed, we found out that already 1 h following skin depilation there is a strong recruitment of neutrophils to the area of application which reaches its maximum at 24 h and is totally resolved 48 h post depilation. The same is also true concerning the integrity of the stratum corneum, as it has been demonstrated by Mittal et al. by measuring the water loss through the skin119. Therefore, the data presented here not only highlight the fact that the inactivated influenza virus can be used for TF vaccination but, to the best of our knowledge this is the first time in which is demonstrated that efficient protective immune responses can be evoked without disruption of the stratum corneum or by breaking the skin barrier by using microneedles, microporation laser devices, or any other physical or chemical treatment of the skin160.

Adjuvants are key molecules in vaccination with their ability to not only enhance antigen-specific immune responses, but also to tailor them according to the specific needs. In addition, since most of the current vaccines favor the stimulation of antigen-specific humoral responses while lacking the activation of the cellular arm of the immune system, such as cytotoxic responses, recent approaches are also addressing the combination of different adjuvants in order to stimulate also Th1 and CTL immune responses, which are

88

able to protect more efficiently against intracellular pathogens like viruses99,100,161. In this regard, we chose the adjuvants c-di-AMP and αGCMPEG for combination with the inactivated H1N1 virus. The c-di-AMP is not only known to stimulate humoral and mixed Th1/Th2/Th17 cellular responses, but also cytotoxic CD8+ T-cells by promoting cross presentation118,134,162,163. Also, αGalCer and its pegylated derivative αGCMPEG have been shown to promote Th1 responses. The αGCMPEG acts via NKT cells leading to the release of cytokines which subsequently results in the stimulation of functionally active NK‐cell subsets164. As expected, the obtained results have shown that although co-administration of either adjuvant only result in a modest increment of H1N1-specific antibody titers, the ratio of IgG1/IgG2c and the cytokine profiles revealed that the Th2 response stimulated by the inactivated H1N1 alone is now strongly shifted toward a more balanced Th1/Th2 response. This is in line with previous observations in which polyphosphazenes were used for immunization by i.n., i.d. and s.c. route83,91,134. In addition, co-administration an adjuvant also enhanced vaccine efficacy, as indicated by a more homogeneous and robust response within the groups, as well as by the stimulation of multifunctional T-cells which play a vital role in vaccine efficacy84,90. Incorporation of c-di-AMP and αGCMPEG did not considerably increase the H1N1-specific antibody titers, as determined by ELISA. However, when the functionality of the stimulated antibody responses was evaluated, it became clear that adjuvantation was critical to enhance the quality of humoral responses (i.e. increased HAI titers in mice receiving the adjuvanted formulations). On the other hand, this can be easily explained by the stimulation of more robust cellular responses, which encompassed Th1 cells which are critical for the induction of the IgG2c antibodies known for the clearance of the virus and increased protection against lethal influenza challenge165,166. We also wondered whether adjuvantation would allow the reduction of the antigen dose. For example, Sadati et al. have demonstrated that intradermal vaccination of mice with a low dose of inactivated H1N1 antigen (0.5 µg) adjuvanted with chitosan or CpG adjuvants resulted in the same range of humoral and cellular immune responses and protection than the non-adjuvanted high antigen dose (5 µg)167. In addition, considering the encouraging data obtained using the epidemic H1N1 virus for TF vaccination, we also wanted to evaluate if TF vaccination would be also useful to elicit strong immune responses against pandemic influenza strains, such as the H5N1 virus. Avian influenza viruses like the H5N1 and H7N9 viruses have shown the ability to infect humans causing severe diseases associated with a higher death rate168–170. Therefore, we performed a vaccination experiment immunizing mice with 2 µg of inactivated H5N1 virus co-administered with either c-di-AMP or its derivative 2’3’-c-di-AM(PS)2. The sulfonated c-di-AMP displays protective features against the enzymatic degradation by phosphodiesterases, thereby

89

increasing its persistence within the body which in turn could increase vaccine efficacy126,171. In line with the results obtained by Sadati et al. both adjuvanted formulations stimulated a similar humoral immune response than the one observed when using the higher dose of inactivated H1N1 virus. The same is true concerning the stimulated H5N1-specific cellular responses. Nevertheless, although not statistically significant, 2’3’-c-di-AM(PS)2 seems to be slightly more efficient in increasing the strength and quality of the stimulated cellular responses underlining our hypothesis that persistence of the adjuvant in the skin could result in enhanced immune activation.

Recent vaccination approaches focus on the so-called subunit vaccines that are made of pathogen subcellular components with well-defined properties. This, in turn, leads to well-defined products which exhibit a better safety profile. Unfortunately, this increased safety profile is achieved by a reduction in the immunogenicity, making necessary the development of effective delivery systems and adjuvants172. In this regard, recent studies have demonstrated the potential of polymeric nanoparticles as antigen delivery system for TF vaccination. They also showed that incorporation of an adjuvant in such vaccine formulation is essential to generate efficient humoral and cellular immune responses, especially when the skin barrier is kept intact119,120. However, so far no vaccination efficacy could be demonstrated using NPs, such as PLGA or lecithin based micellar sugar glass nanoparticles (IMSG)119,120. Because of this, we investigated the usefulness of PCPP and PCEP, two members of the well-known polyphosphazene polymer family, for TF vaccination. Both, PCPP and PCEP have been shown to increase antigen-specific immune responses when used for mucosal and s.c. vaccination strategies83,89,91,135. Thus, aiming at the development of a non-invasive TF vaccine targeting the HF and keeping intact the stratum corneum, we first investigated the penetration capacity of the polyphosphazenes across the skin into the HF. Using in vivo two-photon microscope technology, we could confirm that the NPs are accumulating around the hair and penetrate in the infundibulum of the HF. Thus, polyphosphazenes in our experimental setting showed similar penetration properties as the chitosan-PLGA NPs used for TF immunization resulting in strong antigen-specific immune responses119,120. Our observations are in line with Rancan and co-workers demonstrating that NPs encompassing HIV-1 p24 protein aggregate in the duct of the infundibulum of the HF where they subsequently release the protein antigen. The protein then diffuses in the epidermis and dermis where it is uptaked by the skin DC, especially the LC173.

90

Consistently, characterization of the adaptive immune responses stimulated following TF vaccination of mice with OVA-PCPP and OVA-PCEP co-administered with c-di-AMP resulted in strong antigen-specific antibody production. Moreover, the detected IgG titers were further increased compared to those obtained using OVA-CS/PLGA or IMSG NP co-formulated with c-di-AMP119,120. In line with what has been described by Schulze et al.using polyphosphazenes for mucosal application, OVA-PCPP and OVA-PCEP co-administered with c-di-AMP stimulated a balanced Th1/Th2/Th17 response134. Similar observations have been also reported by others showing mixed Th1/Th2 responses stimulated by PCEP and PCPP84,91,135. However, incorporation of c-di-AMP as adjuvant seems to further improve the modulation of the T helper response towards a more balanced Th1/Th2 response, by promoting a Th1 environment. On the other hand, antigen-specific cellular immune responses could not be further improved using polyphosphazenes as compared to OVA-CS/PLGA NPs117. Thus, neither the quality of OVA-specific T-cell responses nor the strength of the cellular responses have been increased following TF vaccination of mice with both polyphosphazene formulations117. Furthermore, like for the results obtained using OVA-CS/PLGA NPs, also polyphosphazenes stimulate only very modest numbers of OVA-specific cytotoxic T-lymphocytes. This is in contrast to what has been observed when polyphosphazene based OVA formulations were used by the i.n. route, since strong CTL responses were observed in mice receiving OVA-PCEP or OVA-PCPP co-administered with c-di-AMP134. The inability of this formulations to stimulate strong OVA-specific CD8 T-cells when applied by TF route might be one of the reasons why immunized mice were not protected against a sub-lethal challenge with the influenza virus expressing the OVA peptide SIINFEKL which encompasses the TCR receptor recognized by CD8+ T-lymphocytes. Consequently protection against virus infection could have been mediated only via the cellular immune response and not by the humoral immune response as it most properly was the case for the mucosal setting reported by Schulze et al134. Especially in case of influenza infection, it has been reported that HA-specific antibodies play a major role in the resistance against the virus. Indeed, the HA protein represents 80% of the proteins that protrude from the viral membrane174,175, making it a good target for opsonization. Moreover, neutralizing antibodies against the globular head of the HA protein are able to neutralize the virus attachment and entry in the host cells, which make them strong mediators of the resistance to the influenza virus and are considered as an immune correlate of protection175–177. However, recent studies performed in humans have demonstrated that a pre-existing influenza-specific T-cell immune response also correlates with protection in persons lacking neutralizing antibodies142,143,178 179.

91

Accordingly, we then raised the question if protective immune responses can be obtained using polyphosphazene polymers co-formulated with recombinant HA protein of influenza. To this end, mice were immunized following the same experimental setting using HA-PCPP and HA-PCEP, respectively. In addition to c-di-AMP, we also incorporated αGCMPEG as adjuvant. The αGCMPEG was shown to also stimulate mixed Th1/Th2 responses acting via NKT cells180. Admittedly, the frequency of NKT cells in skin seems to be quite low (∼0.03% of cells in healthy skin and ∼0.6% of cells in hyperplastic skin181) but CD1d, the receptor for αGCMPEG, is also expressed by B cells, LC and keratinocytes182,183. Bonish et al. demonstrated that keratinocytes expressing CD1d subsequently can activate NK-T cells to produce IFN-γ, whereas Otsuka and co-workers showed that LC activated by αGCMPEG produce pro-inflammatory cytokines184,185. Therefore, incorporation of αGCMPEG could result in the stimulation of all arms of the immune system, thereby increasing vaccine efficacy. When the stimulated HA-specific immune responses were compared with those obtained using OVA as antigen, lower HA-specific IgG titers were observed regardless of the adjuvant used. However, the overall type of immune response was similar, as indicated by the observed mixed Th1/Th2 responses. Interestingly, while TF immunization with HA-PCPP with or without adjuvantation stimulated an increased IFN production by antigen-specific immune cells, HA-PCEP stimulated enhanced numbers of IFN producing cells. These data are in line with what we have already observed using OVA as antigen, as well as with previous studies using polyphosphazene polymers for vaccination via intranasal and subcutaneous route134,186. While co-administration of αGCMPEG with HA-PCEP resulted in reduced IL-17 levels, this was not the case for the HA-PCPP formulations. This could indicate that the IL-17 detected upon PCEP treatment was produced mainly by Th17 cells, whereas PCPP predominantly stimulates innate immune cells to produce IL-17. In fact, αGCMPEG was shown to inhibit Th17 responses while enhancing IL-17 production by innate immune cells83,86,137,187. Similar levels of bi- and trifunctional T-cells have been elicited following TF vaccination with OVA-PCEP and OVA-PCPP co-administered with c-di-AMP, as compared to those reported by Schulze et al. following an i.n. vaccination strategy134.Remarkably, using recombinant HA as antigen in combination with PCPP or PCEP and c-di-AMP promoted only a weak stimulation of multifunctional CD4+ T-cells. Nevertheless, these results are in line with the observation of Madhun et al. showing that i.m. immunization of mice using plant-derived recombinant HA protein in combination with the cyclic di-nucleotide c-di-GMP stimulated only very low levels of multifunctional CD4+ T-cells, if any at all, even though a 30 times higher dose of HA protein was used, whereas high frequencies were detected following i.n. administration188. Taken together,

92

TF immunization of mice using recombinant HA formulated with either of the two polyphosphazenes PCEP and PCPP co-administered with c-di-AMP or αGCMPEG stimulated increased humoral and cellular immune responses. The differences observed compared to the responses obtained using OVA as antigen could be explained, to some extent, by the different amount of antigen used. The HA-based formulations encompassed 4 times less antigen with respect to those containing OVA as antigen. The vaccine antigen concentration was reduced since the initial experiments using inactivated H5N1 virus revealed the possibility of dose sparing.

A viral infection experiment following one prime and three homologous boosts with HA-PCEP alone or adjuvanted with either c-di-AMP or αGCMPEG was performed to validate vaccine efficacy. The PCEP based formulations were chosen, since previous studies reported protective immune responses following transdermal vaccination using PCEP83,89,189. Consistent with our hypothesis, we observed that mice vaccinated with the adjuvanted formulations were protected against virus infection, as indicated by a strong decrease of viral load as well as lack of deterioration of the physical conditions. In contrast, the non-adjuvanted formulation could protect only a third of the immunized mice.

In conclusion, the data presented here have demonstrated that non-invasive vaccination by the TF route is able to stimulate protective immune responses against infection. Protection can be evoked even by using subunit vaccines, when the vaccine antigen is delivered using a proper delivery system and adjuvant.

In order to generate the basic knowledge to further improve vaccine efficacy, we unraveled the underlying mechanisms resulting in immune activation following TF vaccination with the adjuvanted polyphosphazene based vaccine formulations. Understanding the immune events occurring during TF vaccination would enable to tailor the vaccination strategy to maximize efficacy, minimizing the require antigen dose. To this end, we investigated in detail the first immune events stimulated at the inductive site (i.e. skin) after TF immunization. In line with what has been published, the loading of the vaccine antigen in PCPP nanoparticle formulations increased antigen stability, as well as its delivery to and uptake by DCs190,191. Interestingly, the capacity of antigen uptake by skin DC was mainly triggered by the polymer, having the adjuvant rather limited role. However, when using αGCMPEG, antigen processing DC persist longer in the skin as compared to groups treated with DQ-OVA-PCPP and c-di-AMP. Although not statistically significant, we have shown that mice immunized with DQ-OVA-PCPP co-administrated with αGCMPEG showed also a reduced expression of the activation marker CD86 on XCR1+ DCs and, surprisingly, also

93

on the LC populations with respect to the non-adjuvanted formulation. Thus, it seems that αGCMPEG inhibits the activation, and a fortiori the migration, of XCR1+ DCs and LCs. Since it has been shown that LCs do not express the αGalCer receptor CD1d at steady state but only upon activation192,193, the effect of αGCMPEG might be indirectly driven by bystander cells. Since it has been demonstrated earlier that human keratinocytes express CD1d at steady state184, it could be possible that αGCMPEG activates the keratinocytes, thereby resulting in the release of cytokines and chemokines which then influence LC retention. In contrast, it has been reported that XCR1+ DCs constitutively express the MHC like receptor CD1d and that they are the main subpopulation activating iNKT cells via CD1d bound glycolipids, resulting in either Th1- or Th2-associated cytokine responses35,37,194. Therefore, prolonged persistence of XCR1+ DC might be explained by their interaction with iNKT cells upon αGCMPEG binding. On the other hand, the DN cDC subset, that has been shown to induce Th2 polarization42,195–197 seems to be most responsive to αGCMPEG, as indicated by increased migration rates. Taken together, it seems that the polyphosphazene based formulation co-administered with αGCMPEG is able to stimulate different DC subsets, which in turn leads to a mixed Th1/Th2 response.

In contrast, the adjuvant c-di-AMP seems to be more efficient at activating CD11b+ DCs, DN cDCs, and XCR1+ DCs following TF immunization with DQ-OVA-PCPP as compared to αGCMPEG. The c-di-AMP adjuvanted formulation significantly increased the migration of the XCR1+ DCs. This DC subset not only favors the stimulation of a Th1 based immune responses, but it is further characterized by a high ability to cross-present cell-associated and soluble antigen198–200. However, since skin DCs tend to change the expression of some markers (e.g. CD11b)35, their identification after arrival in the draining LN is difficult. While Hayes et al. claimed that the majority of migrating cells found in the draining LNs are not resident skin cells but rather cells recruited to the skin during inflammation201, it is likely that the majority of skin DCs found in the draining LN after immunization of mice with OVA-PCPP + c-di-AMP are derived from the XCR1+ DC subpopulation146. Thus, considering the fact that Hayes and co-workers did not use the CD207 antibody (or equivalent) to phenotype all the skin DC subsets, it is likely that the CD103+ CD11b+ cell subpopulation detected in the draining LN by Hayes et al. is actually represented by the skin resident XCR1+DC subset. Furthermore, this population has been described to migrate from the skin into the deep T-cell zone (TCZ) within 48 h202, which correlates with the kinetics of the XCR1+ DC migration that we observed after TF immunization using OVA-PCPP + c-di-AMP. Moreover, it has been shown that the priming of T-cells in the TCZ of the draining LN leads to a Th1 polarization. Depending

94

on their localization in the draining LN, these lymphocytes can even license LN resident XCR1+ DCs to induce CTL memory203,204. Altogether, these findings match with our study in which mice vaccinated with OVA or HA antigen co-formulated with a polyphosphazene polymer and adjuvanted with c-di-AMP showed a Th1 polarization as well as a limited degree of induction of CTLs. Besides the Th1 polarization also the IL-17 production observed following vaccination using c-di-AMP could be mediated by the XCR1+ and CD11b+ DC subsets. Indeed, although it has been shown that the main skin DC subset able to induce Th17 polarization is the LC population195,205, XCR1+ and CD11b+ DCs have also been recently highlighted as responsible for the induction of differentiation of IL-17-producing CD8+ T-cells (Tc17)206.

In line with the activation levels and migration capacities observed for the different DC subsets following TF immunization with OVA-PCPP plus adjuvant, the amount of proinflammatory chemokines released locally in the skin is higher as compared to that detected in the skin of mice receiving OVA-PCPP alone. Consistent with the data of Thomas et al., we observed high levels of the CCL3 chemokine in the skin following immunization with the αGCMPEG adjuvanted OVA-PCPP formulation, followed by CCL2 and CXCL1207,208. Thomas and co-workers demonstrated that activation of NKT cells via αGCMPEG-CD1d complexes resulted in CD1d restricted T-cell responses. Furthermore, it has been previously shown that one of the major cell types expressing CCL3 are DN cDCs. Also, Ochiai et al. highlighted the fact that this subset of DC is activated upon TSLP stimulation209, which explain their increased migration activity observed after αGCMPEG treatment210. Although this DC subpopulation seems to play an important role in the innate immune response, only few is know about its exact functions. Besides some evidence that they polarize the immune response toward Th2 via secretion of CCL17211, the exact functions still have to be elucidated. It is also important to note that, although CCL3 plays pleiotropic roles via its different receptors (CCR1, CCR3, and CCR5) present on different cell types, one major cell subset bearing receptors for CCL3 is DCs212. Hence, it is not surprising that also co-administration of c-di-AMP resulted in high levels of CCL3. However, in contrast to αGCMPEG, c-di-AMP also stimulates strong production of CCL2, CCL4 and CXCL1, followed by CCL5 and CXCL10. While CCL3, CCL5 and CXCL1 were shown to play a major role in recruitment of neutrophils and Th1 cells, CCL2 seems to mediate DC recruitment, whereas CCL4 and CXCL10 are responsible for NK cell recruitment207,213–216. In line with these findings, we not only observed an increased neutrophil influx following TF application of c-di-AMP adjuvanted DQ-OVA PCPP, but also the stimulation of Th1/Th2/Th17 responses. In this regard, it has been shown that the

95

release of CXCL1 acts as a chemoattractant molecule for neutrophils, which in turn can induce the recruitment of Th1 and Th17 cells217.

Altogether, the results disclosed here reveal that dependent on the vaccine composition, especially the incorporated adjuvant, antigen-specific immune responses can be tailored by targeting specific skin DC subsets. While c-di-AMP seems to target, directly or indirectly, the XCR1+ DC subset, which seems to be responsible for driving the adaptive immune response toward Th1/Th17 polarization, αGCMPEG turned out to somehow counter the activation of the XCR1+ DC subpopulation and rather promote the DN cDC skin subset. The later in turn could explain the more Th2 biased response observed when vaccinating mice with HA-PCPP + αGCMPEG. Nevertheless, although c-di-AMP and αGCMPEG seem to stimulate different skin DC activation pathways both adjuvants were able to induce protection against influenza challenge when using PCEP as vaccine carrier. This point to the fact that different effector mechanisms contribute to confer protection against the influenza virus, and that different vaccine can display efficacy depending on the antigen, the delivery system and the adjuvant included in the formulation.

96

Conclusion and Outlook

The studies performed in this thesis showed the high potential of TF vaccination for the fight against influenza infections. Indeed, we have demonstrated that the antigen currently used in influenza vaccines can also be utilized for the TF vaccination route, resulting in full protection of against influenza infection in a preclinical model. Moreover, we showed that co-administration of an adjuvant with inactivated virus, not only strengthen the immune response, but also tailor antigen-specific adaptive responses. Finally, we demonstrated that TF vaccination efficiently protects mice against viral infection when using a subunit vaccine based on the HA protein. We, therefore confirmed that the non-invasive TF vaccination strategy is a valid alternative for classical vaccination strategies based on i.m. injection. However, considering the high amount of antigen needed, further efforts have to be undertaken in order to increase vaccine efficacy. For example, elucidating more in detail the underlying mechanisms involved in stimulation of efficient immune responses would allow adapting the vaccine formulations allowing further optimization of antigen dose and the immunization schedule. When investigating the very first events in the skin following non-invasive TF vaccination by focusing on the skin DC subsets, we found that although less immunogenic as compared to inactivated virus, adjuvanted polyphosphazene based formulations are able to fine-tune the immune response by activating some specific DC types, which then polarize the adaptive immune response. In this regard, targeting of a certain DC subset by incorporation of a specific ligand and/or adjuvant or by using a corresponding NP would allow fine tailoring of immune responses for the specific needs. To this end, interactions between skin DC and other skin cells need to be addressed more in detail. A recent study from Palgen et al highlighted the fact that the innate myeloid immune response can differ between prime and booster immunizations148. Therefore, future investigations should also address the possibility to modify the immunization schedule in such a way that prime and booster immunization will be performed using different adjuvants in order to stimulate optimal and long-lasting protective immune responses. This will be particularly important considering the actual trend of developing next-generation vaccines against influenza, which are able to stimulate broader cross-protective responses and confer longer-lasting immunity218.

Other aspects of TF vaccination that have not been considered yet are the impact of other adaptive immune cells and/or the skin microflora. Like in the gut, also the skin has an

97

important microflora. However, only little is known about the interaction between the skin immune system and skin microflora and thus, on its potential role in vaccine efficacy. The same is true for immune cells such as γδ T-cells. Both epidermis and dermis are populated by large numbers of γδ T-cells, which were shown to play for example a key role in the response to Mycobacterium infection219–221. Therefore, detailed investigations in this area should be performed in order to possess all the keys needed to develop safe and effective vaccines.

Altogether, this thesis provides critical knowledge for the development of non-invasive TF vaccination strategies, as well as to decipher key aspects of the early immune events triggered by skin DCs after delivery of adjuvanted formulations applied onto intact epidermis that penetrate the skin via HF. Thus, it is expected that this work will facilitate the development of TF vaccines that not only display high efficacy, but also an improved safety profile, easier administration logistics and higher acceptance by the public.

98

References

1. Global Heath Observatory Data Repository,. GHO (2016).

2. Patil, S. U. & Shreffler, W. G. Novel vaccines: Technology and development. J. Allergy

Clin. Immunol. 143, 844–851 (2019).

3. Ahmed, N. & Gottschalk, S. How to design effective vaccines: lessons from an old success

story. Expert Rev Vaccines 8, 543–546 (2009).

4. Bailey, I. Edward Jenner (1749-1823): naturalist, scientist, country doctor, benefactor to

mankind. J Med Biogr 4, 63–70 (1996).

5. Razzell, P. Edward Jenner: The history of a medical myth. Med Hist 9, 216–229 (1965).

6. Riedel, S. Edward Jenner and the history of smallpox and vaccination. Proc (Bayl Univ

Med Cent) 18, 21–25 (2005).

7. Saliou, P. & Girard, M. From Jenner and Pasteur to vaccinology. Therapie 60, 201–204

(2005).

8. Vartak, A. & Sucheck, S. J. Recent Advances in Subunit Vaccine Carriers. Vaccines

(Basel) 4, (2016).

9. Rappuoli, R., Pizza, M., Del Giudice, G. & De Gregorio, E. Vaccines, new opportunities

for a new society. Proc Natl Acad Sci U S A 111, 12288–12293 (2014).

10. Leask, J. Target the fence-sitters. Nature 473, 443–445 (2011).

11. Lamichhane, A., Azegamia, T. & Kiyonoa, H. The mucosal immune system for vaccine

development. Vaccine 32, 6711–6723 (2014).

12. Neutra, M. R. & Kozlowski, P. A. Mucosal vaccines: the promise and the challenge. Nat.

Rev. Immunol. 6, 148–158 (2006).

99

13. Yuki, Y. & Kiyono, H. Mucosal vaccines: novel advances in technology and delivery.

Expert Rev Vaccines 8, 1083–1097 (2009).

14. Zheng, Z., Diaz-Arévalo, D., Guan, H. & Zeng, M. Noninvasive vaccination against

infectious diseases. Hum Vaccin Immunother 14, 1717–1733 (2018).

15. Riese, P., Schulze, K., Ebensen, T., Prochnow, B. & Guzmán, C. A. Vaccine adjuvants:

key tools for innovative vaccine design. Curr Top Med Chem 13, 2562–2580 (2013).

16. Cox, E., Verdonck, F., Vanrompay, D. & Goddeeris, B. Adjuvants modulating mucosal

immune responses or directing systemic responses towards the mucosa. Vet. Res. 37, 511–539

(2006).

17. Riese, P., Sakthivel, P., Trittel, S. & Guzmán, C. A. Intranasal formulations: promising

strategy to deliver vaccines. Expert Opin Drug Deliv 11, 1619–1634 (2014).

18. Kolarsick, P. A. J., Kolarsick, M. A. & Goodwin, C. Anatomy and Physiology of the Skin.

Journal of the Dermatology Nurses’ Association 3, 203 (2011).

19. Kanitakis, J. Anatomy, histology and immunohistochemistry of normal human skin. Eur J

Dermatol 12, 390–399; quiz 400–401 (2002).

20. Kabashima, K., Honda, T., Ginhoux, F. & Egawa, G. The immunological anatomy of the

skin. Nat Rev Immunol 19, 19–30 (2019).

21. Bos, J. D. & Kapsenberg, M. L. The skin immune system Its cellular constituents and their

interactions. Immunology Today 7, 235–240 (1986).

22. Streilein, J. W. Lymphocyte Traffic, T-Cell Malignancies and the Skin. Journal of

Investigative Dermatology 71, 167–171 (1978).

23. Wayne Streilein, J. Skin-Associated Lymphoid Tissues (SALT): Origins and Functions.

Journal of Investigative Dermatology 80, S12–S16 (1983).

24. Malissen, B., Tamoutounour, S. & Henri, S. The origins and functions of dendritic cells

and macrophages in the skin. Nat. Rev. Immunol. 14, 417–428 (2014).

100

25. Nicolas, J.-F. & Guy, B. Intradermal, epidermal and transcutaneous vaccination: from

immunology to clinical practice. Expert Rev Vaccines 7, 1201–1214 (2008).

26. Merad, M. et al. Langerhans cells renew in the skin throughout life under steady-state

conditions. Nat Immunol 3, 1135–1141 (2002).

27. Greter, M. et al. Stroma-derived interleukin-34 controls the development and maintenance

of langerhans cells and the maintenance of microglia. Immunity 37, 1050–1060 (2012).

28. Kubo, A., Nagao, K., Yokouchi, M., Sasaki, H. & Amagai, M. External antigen uptake by

Langerhans cells with reorganization of epidermal tight junction barriers. J. Exp. Med. 206,

2937–2946 (2009).

29. Idoyaga, J. et al. Specialized role of migratory dendritic cells in peripheral tolerance

induction. J Clin Invest 123, 844–854 (2013).

30. Igyártá, B. Z. et al. Skin-Resident Murine Dendritic Cell Subsets Promote Distinct and

Opposing Antigen-specific T Helper Responses. Immunity 35, 260–272 (2011).

31. Matsue, H., Cruz, P. D., Bergstresser, P. R. & Takashima, A. Profiles of cytokine mRNA

expressed by dendritic epidermal T cells in mice. J. Invest. Dermatol. 101, 537–542 (1993).

32. Toulon, A. et al. A role for human skin-resident T cells in wound healing. J. Exp. Med.

206, 743–750 (2009).

33. Pasparakis, M., Haase, I. & Nestle, F. O. Mechanisms regulating skin immunity and

inflammation. Nature Reviews Immunology 14, 289–301 (2014).

34. Bedoui, S. et al. Cross-presentation of viral and self antigens by skin-derived CD103+

dendritic cells. Nat. Immunol. 10, 488–495 (2009).

35. Henri, S. et al. CD207+ CD103+ dermal dendritic cells cross-present keratinocyte-derived

antigens irrespective of the presence of Langerhans cells. J. Exp. Med. 207, 189–206 (2010).

101

36. Edelson, B. T. et al. Peripheral CD103+ dendritic cells form a unified subset

developmentally related to CD8alpha+ conventional dendritic cells. J. Exp. Med. 207, 823–836

(2010).

37. Haniffa, M. et al. Human Tissues Contain CD141hi Cross-Presenting Dendritic Cells with

Functional Homology to Mouse CD103+ Nonlymphoid Dendritic Cells. Immunity 37, 60–73

(2012).

38. Merad, M., Sathe, P., Helft, J., Miller, J. & Mortha, A. The dendritic cell lineage: ontogeny

and function of dendritic cells and their subsets in the steady state and the inflamed setting.

Annu. Rev. Immunol. 31, 563–604 (2013).

39. Gerner, M. Y., Kastenmuller, W., Ifrim, I., Kabat, J. & Germain, R. N. Histo-Cytometry:

in situ multiplex cell phenotyping, quantification, and spatial analysis applied to dendritic cell

subset micro-anatomy in lymph nodes. Immunity 37, 364–376 (2012).

40. Bell, B. D. et al. STAT5 is critical in Dendritic Cells for development of TH2- but not

TH1-dependent Immunity. Nat Immunol 14, 364–371 (2013).

41. Halim, T. Y. F. et al. Group 2 innate lymphoid cells license dendritic cells to potentiate

memory TH2 cell responses. Nat. Immunol. 17, 57–64 (2016).

42. Kashem, S. W., Haniffa, M. & Kaplan, D. H. Antigen-Presenting Cells in the Skin. Annu.

Rev. Immunol. 35, 469–499 (2017).

43. Weniger, B. G. & Papania, M. J. 61 - Alternative vaccine delivery methods. in Vaccines

(Sixth Edition) (eds. Plotkin, S. A., Orenstein, W. A. & Offit, P. A.) 1200–1231 (W.B. Saunders,

2013). doi:10.1016/B978-1-4557-0090-5.00063-X

44. Engelke, L., Winter, G., Hook, S. & Engert, J. Recent insights into cutaneous

immunization: How to vaccinate via the skin. Vaccine 33, 4663–4674 (2015).

45. Mittal, A. et al. Non-invasive delivery of nanoparticles to hair follicles: a perspective for

transcutaneous immunization. Vaccine 31, 3442–3451 (2013).

102

46. Combadière, B. et al. Preferential amplification of CD8 effector-T cells after

transcutaneous application of an inactivated influenza vaccine: a randomized phase I trial. PLoS

ONE 5, e10818 (2010).

47. Baleeiro, R. B. et al. Topical Vaccination with Functionalized Particles Targeting Dendritic

Cells. Journal of Investigative Dermatology 133, 1933–1941 (2013).

48. Fan, H., Lin, Q., Morrissey, G. R. & Khavari, P. A. Immunization via hair follicles by

topical application of naked DNA to normal skin. Nat. Biotechnol. 17, 870–872 (1999).

49. Rancan, F., Afraz, Z., Combadiere, B., Blume-Peytavi, U. & Vogt, A. Hair Follicle

Targeting with Nanoparticles. in Nanotechnology in Dermatology (eds. Nasir, A., Friedman, A.

& Wang, S.) 95–107 (Springer New York, 2013). doi:10.1007/978-1-4614-5034-4_9

50. Lademann, J. et al. Nanoparticles – An efficient carrier for drug delivery into the hair

follicles. European Journal of Pharmaceutics and Biopharmaceutics 66, 159–164 (2007).

51. Patzelt, A. et al. Selective follicular targeting by modification of the particle sizes. J

Control Release 150, 45–48 (2011).

52. Mathes, C. et al. Tight junctions form a barrier in porcine hair follicles. Eur. J. Cell Biol.

95, 89–99 (2016).

53. Combadiere, B. & Liard, C. Transcutaneous and intradermal vaccination. Human Vaccines

7, 811–827 (2011).

54. Glenn, G. M. et al. Transcutaneous immunization: a human vaccine delivery strategy using

a patch. Nat. Med. 6, 1403–1406 (2000).

55. Glenn, G. M. et al. Safety and immunogenicity of an enterotoxigenic Escherichia coli

vaccine patch containing heat-labile toxin: use of skin pretreatment to disrupt the stratum

corneum. Infect. Immun. 75, 2163–2170 (2007).

56. Etchart, N. et al. Safety and efficacy of transcutaneous vaccination using a patch with the

live-attenuated measles vaccine in humans. Vaccine 25, 6891–6899 (2007).

103

57. Frech, S. A. et al. Use of a patch containing heat-labile toxin from Escherichia coli against

travellers’ diarrhoea: a phase II, randomised, double-blind, placebo-controlled field trial. Lancet

371, 2019–2025 (2008).

58. Vogt, A. et al. Transcutaneous Anti-Influenza Vaccination Promotes Both CD4 and CD8

T Cell Immune Responses in Humans. The Journal of Immunology 180, 1482–1489 (2008).

59. Glenn, G. M. et al. Transcutaneous immunization with cholera toxin protects mice against

lethal mucosal toxin challenge. J. Immunol. 161, 3211–3214 (1998).

60. El-Ghorr, A. A., Williams, R. M., Heap, C. & Norval, M. Transcutaneous immunisation

with herpes simplex virus stimulates immunity in mice. FEMS Immunol. Med. Microbiol. 29,

255–261 (2000).

61. Shi, Z. et al. Protection against tetanus by needle-free inoculation of adenovirus-vectored

nasal and epicutaneous vaccines. J. Virol. 75, 11474–11482 (2001).

62. Eyles, J. E. et al. Immunisation against plague by transcutaneous and intradermal

application of subunit antigens. Vaccine 22, 4365–4373 (2004).

63. Belyakov, I. M., Hammond, S. A., Ahlers, J. D., Glenn, G. M. & Berzofsky, J. A.

Transcutaneous immunization induces mucosal CTLs and protective immunity by migration of

primed skin dendritic cells. J. Clin. Invest. 113, 998–1007 (2004).

64. Skountzou, I., Quan, F.-S., Jacob, J., Compans, R. W. & Kang, S.-M. Transcutaneous

immunization with inactivated influenza virus induces protective immune responses. Vaccine

24, 6110–6119 (2006).

65. Stoitzner, P. et al. Tumor immunotherapy by epicutaneous immunization requires

langerhans cells. J. Immunol. 180, 1991–1998 (2008).

66. Hickey, D. K., Aldwell, F. E. & Beagley, K. W. Transcutaneous immunization with a novel

lipid-based adjuvant protects against Chlamydia genital and respiratory infections. Vaccine 27,

6217–6225 (2009).

104

67. Bachmann, M. F. & Jennings, G. T. Vaccine delivery: a matter of size, geometry, kinetics

and molecular patterns. Nat. Rev. Immunol. 10, 787–796 (2010).

68. Benson, R. A. et al. Antigen presentation kinetics control T cell/dendritic cell interactions

and follicular helper T cell generation in vivo. Elife 4, (2015).

69. Foged, C., Brodin, B., Frokjaer, S. & Sundblad, A. Particle size and surface charge affect

particle uptake by human dendritic cells in an in vitro model. Int J Pharm 298, 315–322 (2005).

70. He, C., Hu, Y., Yin, L., Tang, C. & Yin, C. Effects of particle size and surface charge on

cellular uptake and biodistribution of polymeric nanoparticles. Biomaterials 31, 3657–3666

(2010).

71. Manolova, V. et al. Nanoparticles target distinct dendritic cell populations according to

their size. Eur. J. Immunol. 38, 1404–1413 (2008).

72. CSIRO Molecular Science, Bag 10, Clayton South MDC, Vic 3169, Australia &

Gunatillake, P. Biodegradable synthetic polymers for tissue engineering. eCM 5, 1–16 (2003).

73. Gregory, A. E., Titball, R. & Williamson, D. Vaccine delivery using nanoparticles. Front

Cell Infect Microbiol 3, (2013).

74. Schmid, D. et al. T cell-targeting nanoparticles focus delivery of immunotherapy to

improve antitumor immunity. Nat Commun 8, 1–12 (2017).

75. Friedman, A. D., Claypool, S. E. & Liu, R. The Smart Targeting of Nanoparticles. Curr

Pharm Des 19, 6315–6329 (2013).

76. Akagi, T., Baba, M. & Akashi, M. Biodegradable Nanoparticles as Vaccine Adjuvants and

Delivery Systems: Regulation of Immune Responses by Nanoparticle-Based Vaccine. in

Polymers in Nanomedicine (eds. Kunugi, S. & Yamaoka, T.) 31–64 (Springer Berlin

Heidelberg, 2012). doi:10.1007/12_2011_150

105

77. Uto, T. et al. Comparative activity of biodegradable nanoparticles with aluminum

adjuvants: antigen uptake by dendritic cells and induction of immune response in mice.

Immunol. Lett. 140, 36–43 (2011).

78. Pati, R., Shevtsov, M. & Sonawane, A. Nanoparticle Vaccines Against Infectious Diseases.

Front Immunol 9, (2018).

79. Ison, M. G. et al. Current research on respiratory viral infections: Fourth International

Symposium. Antiviral Res. 55, 227–278 (2002).

80. McNeal, M. M., Rae, M. N. & Ward, R. L. Effects of different adjuvants on rotavirus

antibody responses and protection in mice following intramuscular immunization with

inactivated rotavirus. Vaccine 17, 1573–1580 (1999).

81. Andrianov, A. K., Marin, A. & Chen, J. Synthesis, Properties, and Biological Activity of

Poly[di(sodium carboxylatoethylphenoxy)phosphazene]. Biomacromolecules 7, 394–399

(2006).

82. Mutwiri, G., Benjamin, P., Soita, H. & Babiuk, L. A. Co-administration of

polyphosphazenes with CpG oligodeoxynucleotides strongly enhances immune responses in

mice immunized with Hepatitis B virus surface antigen. Vaccine 26, 2680–2688 (2008).

83. Garlapati, S. et al. Immunization with PCEP microparticles containing pertussis toxoid,

CpG ODN and a synthetic innate defense regulator peptide induces protective immunity against

pertussis. Vaccine 29, 6540–6548 (2011).

84. Dar, A. et al. Administration of poly[di(sodium carboxylatoethylphenoxy)]phosphazene

(PCEP) as adjuvant activated mixed Th1/Th2 immune responses in pigs. Veterinary

Immunology and Immunopathology 146, 289–295 (2012).

85. Lakshmi, S., Katti, D. S. & Laurencin, C. T. Biodegradable polyphosphazenes for drug

delivery applications. Adv. Drug Deliv. Rev. 55, 467–482 (2003).

106

86. Magiri, R., Mutwiri, G. & Wilson, H. L. Recent advances in experimental

polyphosphazene adjuvants and their mechanisms of action. Cell Tissue Res. 374, 465–471

(2018).

87. Eng, N. F. et al. The potential of polyphosphazenes for delivery of vaccine antigens and

immunotherapeutic agents. Curr Drug Deliv 7, 13–20 (2010).

88. Andrianov, A. K. et al. PCPP-Formulated H5N1 Influenza Vaccine Displays Improved

Stability and Dose-Sparing Effect in Lethal Challenge Studies. Journal of Pharmaceutical

Sciences 100, 1436–1443 (2011).

89. Eng, N. F., Garlapati, S., Gerdts, V., Babiuk, L. A. & Mutwiri, G. K. PCEP enhances IgA

mucosal immune responses in mice following different immunization routes with influenza

virus antigens. Journal of Immune Based Therapies and Vaccines 8, 4 (2010).

90. Magiri, R. et al. Intradermal immunization with inactivated swine influenza virus and

adjuvant polydi(sodium carboxylatoethylphenoxy)phosphazene (PCEP) induced humoral and

cell-mediated immunity and reduced lung viral titres in pigs. Vaccine 36, 1606–1613 (2018).

91. Mutwiri, G. et al. Poly[di(sodium carboxylatoethylphenoxy)phosphazene] (PCEP) is a

potent enhancer of mixed Th1/Th2 immune responses in mice immunized with influenza virus

antigens. Vaccine 25, 1204–1213 (2007).

92. Andrianov, A. K. et al. Poly[di(carboxylatophenoxy)phosphazene] is a potent adjuvant for

intradermal immunization. Proc. Natl. Acad. Sci. U.S.A. 106, 18936–18941 (2009).

93. Ramon, G. & Ramón, G. Sur l’augmentation anormale de l’antitoxine chez les chevaux

producteurs de serum antidiphterique. (1925).

94. Ramon G. Procedes pour accroitre la production des antitoxins. Ann. Inst. Pasteur 40,

(1926).

95. Glenny AT, Pope CG, Waddington H, Wallace V. The Antigenic Value of Toxoid

Precipitated by Potassium Alum. J, Path. Bacteriol. 29, 38–45 (1926).

107

96. Freund, J., Casals, J. & Hosmer, E. P. Sensitization and Antibody Formation after Injection

of Tubercle Bacilli and Paraffin Oil. Proceedings of the Society for Experimental Biology and

Medicine 37, 509–513 (1937).

97. Petrovsky, N. & Aguilar, J. C. Vaccine adjuvants: current state and future trends. Immunol.

Cell Biol. 82, 488–496 (2004).

98. McElrath, M. J. Selection of potent immunological adjuvants for vaccine construction.

Semin. Cancer Biol. 6, 375–385 (1995).

99. Ebensen, T., Delandre, S., Prochnow, B., Guzmán, C. A. & Schulze, K. The Combination

Vaccine Adjuvant System Alum/c-di-AMP Results in Quantitative and Qualitative Enhanced

Immune Responses Post Immunization. Front Cell Infect Microbiol 9, (2019).

100. Garçon, N. & Di Pasquale, A. From discovery to licensure, the Adjuvant System story.

Hum Vaccin Immunother 13, 19–33 (2017).

101. Cheung, T. K. W. & Poon, L. L. M. Biology of Influenza A Virus. Annals of the New York

Academy of Sciences 1102, 1–25 (2007).

102. Akira, S. & Takeda, K. Toll-like receptor signalling. Nat Rev Immunol 4, 499–511 (2004).

103. Kopp, E. B. & Medzhitov, R. The Toll-receptor family and control of innate immunity.

Current Opinion in Immunology 11, 13–18 (1999).

104. Kawai, T. & Akira, S. TLR signaling. Seminars in Immunology 19, 24–32 (2007).

105. Beutler, B. & Rietschel, E. T. Innate immune sensing and its roots: the story of endotoxin.

Nat Rev Immunol 3, 169–176 (2003).

106. Iwasaki, A. & Pillai, P. S. Innate immunity to influenza virus infection. Nat. Rev. Immunol.

14, 315–328 (2014).

107. Kreijtz, J. H. C. M., Fouchier, R. a. M. & Rimmelzwaan, G. F. Immune responses to

influenza virus infection. Virus Res. 162, 19–30 (2011).

108

108. Braciale, T. J., Sun, J. & Kim, T. S. Regulating the adaptive immune response to respiratory

virus infection. Nat. Rev. Immunol. 12, 295–305 (2012).

109. WHO guidelines. (2018).

110. Molinari, N.-A. M. et al. The annual impact of seasonal influenza in the US: measuring

disease burden and costs. Vaccine 25, 5086–5096 (2007).

111. Coelingh, K. L., Luke, C. J., Jin, H. & Talaat, K. R. Development of live attenuated

influenza vaccines against pandemic influenza strains. Expert Rev Vaccines 13, 855–871

(2014).

112. Arvas, A. Vaccination in patients with immunosuppression. Turk Pediatri Ars 49, 181–185

(2014).

113. Cox, M. M. J., Izikson, R., Post, P. & Dunkle, L. Safety, efficacy, and immunogenicity of

Flublok in the prevention of seasonal influenza in adults. Ther Adv Vaccines 3, 97–108 (2015).

114. Osterholm, M. T., Kelley, N. S., Sommer, A. & Belongia, E. A. Efficacy and effectiveness

of influenza vaccines: a systematic review and meta-analysis. Lancet Infect Dis 12, 36–44

(2012).

115. Ohmit, S. E. et al. Influenza vaccine effectiveness in the community and the household.

Clin. Infect. Dis. 56, 1363–1369 (2013).

116. Russell, K. et al. Influenza vaccine effectiveness in older adults compared with younger

adults over five seasons. Vaccine 36, 1272–1278 (2018).

117. Grohskopf, L. A. et al. Prevention and control of seasonal influenza with vaccines:

recommendations of the Advisory Committee on Immunization Practices (ACIP) -- United

States, 2014-15 influenza season. MMWR Morb. Mortal. Wkly. Rep. 63, 691–697 (2014).

118. Ebensen, T., Schulze, K., Riese, P., Morr, M. & Guzmán, C. A. The bacterial second

messenger cdiGMP exhibits promising activity as a mucosal adjuvant. Clin. Vaccine Immunol.

14, 952–958 (2007).

109

119. Mittal, A. et al. Efficient nanoparticle-mediated needle-free transcutaneous vaccination via

hair follicles requires adjuvantation. Nanomedicine: Nanotechnology, Biology and Medicine 11,

147–154 (2015).

120. Mittal, A. et al. Inverse micellar sugar glass (IMSG) nanoparticles for transfollicular

vaccination. Journal of Controlled Release 206, 140–152 (2015).

121. Danilchanka, O. & Mekalanos, J. J. Cyclic dinucleotides and the innate immune response.

Cell 154, 962–970 (2013).

122. Witte, G., Hartung, S., Büttner, K. & Hopfner, K.-P. Structural biochemistry of a bacterial

checkpoint protein reveals diadenylate cyclase activity regulated by DNA recombination

intermediates. Mol. Cell 30, 167–178 (2008).

123. Parvatiyar, K. et al. DDX41 recognizes bacterial secondary messengers cyclic di-GMP and

cyclic di-AMP to activate a type I interferon immune response. Nat Immunol 13, 1155–1161

(2012).

124. Kawano, T. et al. CD1d-Restricted and TCR-Mediated Activation of Vα14 NKT Cells by

Glycosylceramides. Science 278, 1626–1629 (1997).

125. Singh, A. K. et al. Natural Killer T Cell Activation Protects Mice Against Experimental

Autoimmune Encephalomyelitis. Journal of Experimental Medicine 194, 1801–1811 (2001).

126. Yan, H., Wang, X., KuoLee, R. & Chen, W. Synthesis and immunostimulatory properties

of the phosphorothioate analogues of cdiGMP. Bioorg. Med. Chem. Lett. 18, 5631–5634 (2008).

127. Lenz, N. et al. Antiviral Innate Immune Activation in HIV-Infected Adults Negatively

Affects H1/IC31-Induced Vaccine-Specific Memory CD4+ T Cells. Clinical and vaccine

immunology : CVI 22, 688–696 (2015).

128. Seder, R. A., Darrah, P. A. & Roederer, M. T-cell quality in memory and protection:

implications for vaccine design. Nat. Rev. Immunol. 8, 247–258 (2008).

110

129. Derrick, S. C., Yabe, I. M., Yang, A. & Morris, S. L. Vaccine-induced anti-tuberculosis

protective immunity in mice correlates with the magnitude and quality of multifunctional CD4

T cells. Vaccine 29, 2902–2909 (2011).

130. Thakur, A., Pedersen, L. E. & Jungersen, G. Immune markers and correlates of protection

for vaccine induced immune responses. Vaccine 30, 4907–4920 (2012).

131. Moretto, M. M., Harrow, D. I. & Khan, I. A. Effector CD8 T cell immunity in

microsporidial infection: a lone defense mechanism. Semin Immunopathol 37, 281–287 (2015).

132. Vogt, A. et al. Hair follicle targeting, penetration enhancement and Langerhans cell

activation make cyanoacrylate skin surface stripping a promising delivery technique for

transcutaneous immunization with large molecules and particle-based vaccines. Exp. Dermatol.

24, 73–75 (2015).

133. Rancan, F. et al. Topically applied virus-like particles containing HIV-1 Pr55gag protein

reach skin antigen-presenting cells after mild skin barrier disruption. J Control Release 268,

296–304 (2017).

134. Schulze, K., Ebensen, T., Babiuk, L. A., Gerdts, V. & Guzman, C. A. Intranasal vaccination

with an adjuvanted polyphosphazenes nanoparticle-based vaccine formulation stimulates

protective immune responses in mice. Nanomedicine 13, 2169–2178 (2017).

135. Shim, D.-H. et al. Efficacy of poly[di(sodium carboxylatophenoxy)phosphazene] (PCPP)

as mucosal adjuvant to induce protective immunity against respiratory pathogens. Vaccine 28,

2311–2317 (2010).

136. Mars, L. T. et al. Invariant NKT cells inhibit development of the Th17 lineage. PNAS 106,

6238–6243 (2009).

137. Zygmunt, B. M., Weissmann, S. F. & Guzman, C. A. NKT cell stimulation with α-

galactosylceramide results in a block of Th17 differentiation after intranasal immunization in

mice. PLoS ONE 7, e30382 (2012).

111

138. Chen, X. et al. Host Immune Response to Influenza A Virus Infection. Front Immunol 9,

(2018).

139. Brown, D. M., Lee, S., Garcia-Hernandez, M. de la L. & Swain, S. L. Multifunctional CD4

cells expressing gamma interferon and perforin mediate protection against lethal influenza virus

infection. J. Virol. 86, 6792–6803 (2012).

140. Swain, S. L., McKinstry, K. K. & Strutt, T. M. Expanding roles for CD4+ T cells in

immunity to viruses. Nat. Rev. Immunol. 12, 136–148 (2012).

141. Grant, E. J., Quiñones-Parra, S. M., Clemens, E. B. & Kedzierska, K. Human influenza

viruses and CD8(+) T cell responses. Curr Opin Virol 16, 132–142 (2016).

142. Wilkinson, T. M. et al. Preexisting influenza-specific CD4+ T cells correlate with disease

protection against influenza challenge in humans. Nat. Med. 18, 274–280 (2012).

143. McMichael, A. J., Gotch, F. M., Noble, G. R. & Beare, P. A. S. Cytotoxic T-Cell Immunity

to Influenza. New England Journal of Medicine 309, 13–17 (1983).

144. Czakó, R. et al. In Vivo Imaging of Influenza Virus Infection in Immunized Mice. mBio 8,

e00714-17 (2017).

145. Guilliams, M. et al. Skin-draining lymph nodes contain dermis-derived CD103− dendritic

cells that constitutively produce retinoic acid and induce Foxp3+ regulatory T cells. Blood 115,

1958–1968 (2010).

146. Tomura, M. et al. Tracking and quantification of dendritic cell migration and antigen

trafficking between the skin and lymph nodes. Scientific Reports 4, 6030 (2014).

147. Di Pilato, M. & Esteban, M. Neutrophil and vaccine. Cell Cycle 14, 1615–1616 (2015).

148. Palgen, J.-L. et al. Prime and Boost Vaccination Elicit a Distinct Innate Myeloid Cell

Immune Response. Sci Rep 8, (2018).

149. Duffy, D. et al. Neutrophils transport antigen from the dermis to the bone marrow,

initiating a source of memory CD8+ T cells. Immunity 37, 917–929 (2012).

112

150. Hajj Hussein, I. et al. Vaccines Through Centuries: Major Cornerstones of Global Health.

Front Public Health 3, (2015).

151. Lewnard, J. A. & Cobey, S. Immune History and Influenza Vaccine Effectiveness.

Vaccines (Basel) 6, (2018).

152. Belongia, E. A. & McLean, H. Q. Influenza Vaccine Effectiveness: Defining the H3N2

Problem. Clin. Infect. Dis. (2019). doi:10.1093/cid/ciz411

153. Geiß, Y. & Dietrich, U. Catch Me If You Can--The Race Between HIV and Neutralizing

Antibodies. AIDS Rev 17, 107–113 (2015).

154. Paules, C. I., Marston, H. D. & Fauci, A. S. Measles in 2019 — Going Backward. New

England Journal of Medicine (2019). doi:10.1056/NEJMp1905099

155. Measles cases spike globally due to gaps in vaccination coverage. Available at:

https://www.who.int/news-room/detail/29-11-2018-measles-cases-spike-globally-due-to-gaps-

in-vaccination-coverage. (Accessed: 25th August 2019)

156. Dabbagh, A. et al. Progress towards regional measles elimination – worldwide, 2000–

2017. WEEKLY EPIDEMIOLOGICAL RECORD 12 (2018).

157. Lademann, J. et al. Penetration and storage of particles in human skin: perspectives and

safety aspects. Eur J Pharm Biopharm 77, 465–468 (2011).

158. Liard, C. et al. Targeting of HIV-p24 particle-based vaccine into differential skin layers

induces distinct arms of the immune responses. Vaccine 29, 6379–6391 (2011).

159. Mahe, B. et al. Nanoparticle-based targeting of vaccine compounds to skin antigen-

presenting cells by hair follicles and their transport in mice. J. Invest. Dermatol. 129, 1156–

1164 (2009).

160. Garg, S. et al. Needle-Free Skin Patch Delivery of a Vaccine for a Potentially Pandemic

Influenza Virus Provides Protection against Lethal Challenge in Mice. Clin Vaccine Immunol

14, 926–928 (2007).

113

161. Di Pasquale, A. et al. Vaccine safety evaluation: Practical aspects in assessing benefits and

risks. Vaccine 34, 6672–6680 (2016).

162. Ebensen, T. et al. Mucosal Administration of Cycle-Di-Nucleotide-Adjuvanted Virosomes

Efficiently Induces Protection against Influenza H5N1 in Mice. Front Immunol 8, 1223 (2017).

163. Ebensen, T. et al. The bacterial second messenger cyclic diGMP exhibits potent adjuvant

properties. Vaccine 25, 1464–1469 (2007).

164. Riese, P. et al. Activated NKT cells imprint NK-cell differentiation, functionality and

education. Eur. J. Immunol. 45, 1794–1807 (2015).

165. Huber, V. C. et al. Distinct contributions of vaccine-induced immunoglobulin G1 (IgG1)

and IgG2a antibodies to protective immunity against influenza. Clin. Vaccine Immunol. 13,

981–990 (2006).

166. Grødeland, G., Fossum, E. & Bogen, B. Polarizing T and B Cell Responses by APC-

Targeted Subunit Vaccines. Front Immunol 6, 367 (2015).

167. Sadati, S. F. et al. Simultaneous formulation of influenza vaccine and chitosan

nanoparticles within CpG oligodesoxi nucleotides leads to dose-sparing and protects against

lethal challenge in the mouse model. Pathog Dis 76, (2018).

168. Wang, T. T., Parides, M. K. & Palese, P. Seroevidence for H5N1 Influenza Infections in

Humans: Meta-Analysis. Science 335, 1463–1463 (2012).

169. Morens, D. M. & Taubenberger, J. K. How low is the risk of influenza A(H5N1) infection?

J. Infect. Dis. 211, 1364–1366 (2015).

170. Koutsakos, M., Kedzierska, K. & Subbarao, K. Immune Responses to Avian Influenza

Viruses. The Journal of Immunology 202, 382–391 (2019).

171. Foote, J. B. et al. A STING Agonist Given with OX40 Receptor and PD-L1 Modulators

Primes Immunity and Reduces Tumor Growth in Tolerized Mice. Cancer Immunol Res 5, 468–

479 (2017).

114

172. Ebensen, T. & Guzmán, C. A. Immune modulators with defined molecular targets:

cornerstone to optimize rational vaccine design. Hum Vaccin 4, 13–22 (2008).

173. Rancan, F. et al. Particle-based transcutaneous administration of HIV-1 p24 protein to

human skin explants and targeting of epidermal antigen presenting cells. Journal of Controlled

Release 176, 115–122 (2014).

174. Wrigley, N. G. ELECTRON MICROSCOPY OF INFLUENZA VIRUS. Br Med Bull 35,

35–38 (1979).

175. Hobson, D., Curry, R. L., Beare, A. S. & Ward-Gardner, A. The role of serum

haemagglutination-inhibiting antibody in protection against challenge infection with influenza

A2 and B viruses. J Hyg (Lond) 70, 767–777 (1972).

176. Reber, A. & Katz, J. Immunological assessment of influenza vaccines and immune

correlates of protection. Expert Review of Vaccines 12, 519–536 (2013).

177. Potter, C. W. & Oxford, J. S. Determinants of immunity to influenza infection in man. Br.

Med. Bull. 35, 69–75 (1979).

178. Altenburg, A. F., Rimmelzwaan, G. F. & de Vries, R. D. Virus-specific T cells as correlate

of (cross-)protective immunity against influenza. Vaccine 33, 500–506 (2015).

179. McElhaney, J. E. et al. T cell responses are better correlates of vaccine protection in the

elderly. J. Immunol. 176, 6333–6339 (2006).

180. Godfrey, D. I. & Kronenberg, M. Going both ways: immune regulation via CD1d-

dependent NKT cells. J. Clin. Invest. 114, 1379–1388 (2004).

181. Mattarollo, S. R. et al. Invariant NKT cells in hyperplastic skin induce a local immune

suppressive environment by IFN-gamma production. J. Immunol. 184, 1242–1250 (2010).

182. McKee, S. J., Mattarollo, S. R. & Leggatt, G. R. Immunosuppressive roles of natural killer

T (NKT) cells in the skin. J. Leukoc. Biol. 96, 49–54 (2014).

115

183. Lang, G. A., Devera, T. S. & Lang, M. L. Requirement for CD1d expression by B cells to

stimulate NKT cell-enhanced antibody production. Blood 111, 2158–2162 (2008).

184. Bonish, B. et al. Overexpression of CD1d by Keratinocytes in Psoriasis and CD1d-

Dependent IFN-γ Production by NK-T Cells. The Journal of Immunology 165, 4076–4085

(2000).

185. Otsuka, Y. et al. Differentiation of Langerhans Cells from Monocytes and Their Specific

Function in Inducing IL-22-Specific Th Cells. J. Immunol. 201, 3006–3016 (2018).

186. Eng, N. F., Garlapati, S., Lai, K., Gerdts, V. & Mutwiri, G. K. Polyphosphazenes Enhance

Mucosal and Systemic Immune Responses in Mice Immunized Intranasally with Influenza

Antigens. in (2009). doi:10.2174/1875035400902010134

187. Yu, J.-S. et al. Differentiation of IL-17-Producing Invariant Natural Killer T Cells Requires

Expression of the Transcription Factor c-Maf. Front Immunol 8, 1399 (2017).

188. Madhun, A. S. et al. Intranasal c-di-GMP-adjuvanted plant-derived H5 influenza vaccine

induces multifunctional Th1 CD4+ cells and strong mucosal and systemic antibody responses

in mice. Vaccine 29, 4973–4982 (2011).

189. Magiri, R. et al. Intradermal immunization with inactivated swine influenza virus and

adjuvant polydi(sodium carboxylatoethylphenoxy)phosphazene (PCEP) induced humoral and

cell-mediated immunity and reduced lung viral titres in pigs. Vaccine 36, 1606–1613 (2018).

190. Palmer, C. D. et al. The effect of stable macromolecular complexes of ionic

polyphosphazene on HIV Gag antigen and on activation of human dendritic cells and

presentation to T-cells. Biomaterials 35, 8876–8886 (2014).

191. Andrianov, A. K., Marin, A. & Roberts, B. E. Polyphosphazene Polyelectrolytes:  A Link

between the Formation of Noncovalent Complexes with Antigenic Proteins and

Immunostimulating Activity. Biomacromolecules 6, 1375–1379 (2005).

192. Fukunaga, A. et al. Langerhans Cells Serve as Immunoregulatory Cells by Activating NKT

Cells. J Immunol 185, 4633–4640 (2010).

116

193. Gerlini, G. et al. CD1d is Expressed on Dermal Dendritic Cells and Monocyte-Derived

Dendritic Cells. Journal of Investigative Dermatology 117, 576–582 (2001).

194. Arora, P. et al. A single subset of dendritic cells controls the cytokine bias of natural killer

T cell responses to diverse glycolipid antigens. Immunity 40, 105–116 (2014).

195. Kashem, S. W. et al. Candida albicans morphology and dendritic cell subsets determine T

helper cell differentiation. Immunity 42, 356–366 (2015).

196. Martínez-López, M., Iborra, S., Conde-Garrosa, R. & Sancho, D. Batf3-dependent

CD103+ dendritic cells are major producers of IL-12 that drive local Th1 immunity against

Leishmania major infection in mice. Eur. J. Immunol. 45, 119–129 (2015).

197. Bedoui, S. & Heath, W. R. Krüppel-ling of IRF4-Dependent DCs into Two Functionally

Distinct DC Subsets. Immunity 42, 785–787 (2015).

198. Schnorrer, P. et al. The dominant role of CD8+ dendritic cells in cross-presentation is not

dictated by antigen capture. Proc. Natl. Acad. Sci. U.S.A. 103, 10729–10734 (2006).

199. Hildner, K. et al. Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in

cytotoxic T cell immunity. Science 322, 1097–1100 (2008).

200. Pooley, J. L., Heath, W. R. & Shortman, K. Cutting edge: intravenous soluble antigen is

presented to CD4 T cells by CD8- dendritic cells, but cross-presented to CD8 T cells by CD8+

dendritic cells. J. Immunol. 166, 5327–5330 (2001).

201. Hayes, A. J. et al. Spatiotemporal Modeling of the Key Migratory Events During the

Initiation of Adaptive Immunity. Front Immunol 10, 598 (2019).

202. Kitano, M. et al. Imaging of the cross-presenting dendritic cell subsets in the skin-draining

lymph node. Proc. Natl. Acad. Sci. U.S.A. 113, 1044–1049 (2016).

203. Eickhoff, S. et al. Robust Anti-viral Immunity Requires Multiple Distinct T Cell-Dendritic

Cell Interactions. Cell 162, 1322–1337 (2015).

117

204. Hor, J. L. et al. Spatiotemporally Distinct Interactions with Dendritic Cell Subsets

Facilitates CD4+ and CD8+ T Cell Activation to Localized Viral Infection. Immunity 43, 554–

565 (2015).

205. Kobayashi, T. et al. Dysbiosis and Staphylococcus aureus Colonization Drives

Inflammation in Atopic Dermatitis. Immunity 42, 756–766 (2015).

206. Naik, S. et al. Commensal-dendritic-cell interaction specifies a unique protective skin

immune signature. Nature 520, 104–108 (2015).

207. Di Meglio, P., Perera, G. K. & Nestle, F. O. The multitasking organ: recent insights into

skin immune function. Immunity 35, 857–869 (2011).

208. Thomas, S. Y. et al. CD1d-restricted NKT cells express a chemokine receptor profile

indicative of Th1-type inflammatory homing cells. J. Immunol. 171, 2571–2580 (2003).

209. Ochiai, S. et al. CD326loCD103loCD11blo Dermal Dendritic Cells Are Activated by

Thymic Stromal Lymphopoietin during Contact Sensitization in Mice. The Journal of

Immunology 193, 2504–2511 (2014).

210. Proietto, A. I. et al. Differential production of inflammatory chemokines by murine

dendritic cell subsets. Immunobiology 209, 163–172 (2004).

211. Tussiwand, R. et al. Klf4 expression in conventional dendritic cells is required for T helper

2 cell responses. Immunity 42, 916–928 (2015).

212. Viola, A. & Luster, A. D. Chemokines and Their Receptors: Drug Targets in Immunity and

Inflammation. Annu. Rev. Pharmacol. Toxicol. 48, 171–197 (2008).

213. Sawant, K. V. et al. Chemokine CXCL1 mediated neutrophil recruitment: Role of

glycosaminoglycan interactions. Sci Rep 6, 33123 (2016).

214. Xu, L. L., Warren, M. K., Rose, W. L., Gong, W. & Wang, J. M. Human recombinant

monocyte chemotactic protein and other C-C chemokines bind and induce directional migration

of dendritic cells in vitro. J. Leukoc. Biol. 60, 365–371 (1996).

118

215. Bystry, R. S., Aluvihare, V., Welch, K. A., Kallikourdis, M. & Betz, A. G. B cells and

professional APCs recruit regulatory T cells via CCL4. Nat. Immunol. 2, 1126–1132 (2001).

216. Vazirinejad, R., Ahmadi, Z., Kazemi Arababadi, M., Hassanshahi, G. & Kennedy, D. The

biological functions, structure and sources of CXCL10 and its outstanding part in the

pathophysiology of multiple sclerosis. Neuroimmunomodulation 21, 322–330 (2014).

217. Pelletier, M. et al. Evidence for a cross-talk between human neutrophils and Th17 cells.

Blood 115, 335–343 (2010).

218. Khurana, S. Development and Regulation of Novel Influenza Virus Vaccines: A United

States Young Scientist Perspective. Vaccines 6, 24 (2018).

219. Marshall, A. S., Silva, J. R., Bannerman, C. A., Gilron, I. & Ghasemlou, N. Skin-Resident

γδ T Cells Exhibit Site-Specific Morphology and Activation States. J Immunol Res 2019,

(2019).

220. McGill, J. L. et al. The role of gamma delta T cells in immunity to Mycobacterium bovis

infection in cattle. Vet. Immunol. Immunopathol. 159, 133–143 (2014).

221. Lee, J., Choi, K., Olin, M. R., Cho, S.-N. & Molitor, T. W. Gammadelta T cells in immunity

induced by Mycobacterium bovis bacillus Calmette-Guérin vaccination. Infect. Immun. 72,

1504–1511 (2004).