Targeted Nanovaccines Against Respiratory Pathogens Jonathan Tate Goodman Iowa State University

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2016 Targeted nanovaccines against respiratory pathogens Jonathan Tate Goodman Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Allergy and Immunology Commons, Chemical Engineering Commons, Immunology and Infectious Disease Commons, Medical Immunology Commons, and the Polymer Chemistry Commons

Recommended Citation Goodman, Jonathan Tate, "Targeted nanovaccines against respiratory pathogens" (2016). Graduate Theses and Dissertations. 15023. https://lib.dr.iastate.edu/etd/15023

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. i

Targeted nanovaccines against respiratory pathogens

Jonathan Tate Goodman

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Chemical Engineering

Program of Study Committee: Balaji Narasimhan, Major Professor Michael J. Wannemuehler Ian C. Schneider Jacqueline V. Shanks Marian L. Kohut

Iowa State University

Ames, Iowa

2016

TABLE OF CONTENTS ACKNOWLEDGEMENTS ...... viii CHAPTER 1: INTRODUCTION ...... 1 1.1 References ...... 2 1.2 Figures ...... 5 CHAPTER 2: LITERATURE REVIEW ...... 6 2.1 Polymeric Vaccine Delivery Systems ...... 6 2.1.1 Introduction ...... 6 2.1.2 Polyesters ...... 8 2.1.3 Polyanhydrides ...... 12 2.1.4 Natural polymers and polysaccharides ...... 16 2.1.5 Liposomes ...... 18 2.2 Targeted Delivery of Vaccines ...... 20 2.2.1 Introduction ...... 20 2.2.2 Types of PRRs ...... 21 2.2.3 Co-stimulation and crosstalk ...... 28 2.2.4 Targeting mechanisms ...... 29 2.2.5 Targeted vaccines for respiratory viruses ...... 36 2.3 Influenza Virus and Influenza Vaccines ...... 37 2.3.1 Introduction ...... 37 2.3.2 Structure and pathogenicity ...... 38 2.3.3 H1N1 Virus ...... 45 2.3.4 Traditional vaccines and new vaccination approaches ...... 46 2.4 Conclusions ...... 53 2.5 References ...... 54 2.6 Figures ...... 77 2.7 Tables ...... 84 CHAPTER 3: RESEARCH OBJECTIVE AND THESIS ORGANIZATION ...... 88 3.1 Research Objectives ...... 88 3.2 Thesis Organization ...... 90

iii

CHAPTER 4: NANOPARTICLE CHEMISTRY AND FUNCTIONALIZATION DIFFERENTIALLY REGULATES DENDRITIC CELL-NANOPARTICLE INTERACTIONS AND TRIGGERS DENDRITIC CELL MATURATION ...... 91 4.1 Abstract ...... 92 4.2 Introduction ...... 93 4.3. Materials and Methods ...... 95 4.3.1 Materials ...... 95 4.3.2 Monomer and Polymer Synthesis ...... 97 4.3.3 Nanoparticle Synthesis: ...... 97 4.3.4 Nanoparticle Functionalization ...... 98 4.3.5 Characterization of Functionalized Nanoparticles ...... 98 4.3.6 Dendritic Cell Culture and Stimulation...... 99 4.3.7 Flow Cytometric Analysis and Cytokine Secretion Assays ...... 100 4.3.8 Serum Protein Adsorption onto Nanoparticles ...... 100 4.3.9 Characterization of Protein Adsorption ...... 101 4.3.10 Statistical Analysis ...... 102 4.4 Results ...... 102 4.4.1 Nanoparticle characterization ...... 102 4.4.2 Serum protein adsorption influenced 20:80 CPH:SA nanoparticle internalization ...... 103 4.4.3 DC surface marker expression is differentially regulated by chemistry and functionalization ...... 104 4.4.4 DC cytokine production is differentially regulated by nanoparticle chemistry and functionalization ...... 105 4.4.5 Adsorption of immune activating serum proteins onto polyanhydride nanoparticles ...... 106 4.5 Discussion ...... 107 4.6 Conclusions ...... 114 4.7 Acknowledgments ...... 115 4.8 References ...... 115 4.9 Figures ...... 120 4.10 Tables ...... 126 4.11 Supporting Information ...... 128

CHAPTER 5: SAFETY AND BIOCOMPATIBILITY OF CARBOHYDRATE- FUNCTIONALIZED POLYANHYDRIDE NANOPARTICLES ...... 131 5.1 Abstract ...... 132 5.2 Introduction ...... 132 5.3 Materials and Methods ...... 134 5.3.1 Materials ...... 134 5.3.2 Monomer and polymer synthesis ...... 135 5.3.3 Nanoparticle synthesis ...... 136 5.3.4 Sugar synthesis ...... 136 5.3.5 Surface functionalization ...... 137 5.3.6 Nanoparticle characterization ...... 138 5.3.7 Mice ...... 138 5.3.8 Liver and kidney histological and biomarker examination ...... 138 5.3.9 Serum biomarker analysis ...... 139 5.3.10 Urine creatinine and total protein quantification analysis ...... 139 5.3.11 Intranasal administration of particle formulations ...... 139 5.3.12 Lung histological evaluation ...... 140 5.3.13 Flow cytometric analysis of lung tissue ...... 140 5.3.14 Bronchoalveolar lavage (BAL) fluid collection ...... 141 5.3.15 Cytokine and chemokine analysis ...... 141 5.3.16 Statistical analysis ...... 142 5.4 Results ...... 142 5.4.1 Functionalization and characterization of carbohydrate-modified nanoparticles ...... 142 5.4.2 Kidney histological evaluation and renal function ...... 143 5.4.3 Liver histological evaluation and hepatic function ...... 144 5.4.4 Lung histological evaluation ...... 145 5.4.5 Distribution of lung cellular populations ...... 146 5.4.6 Cytokine/chemokine secretion ...... 146 5.5 Discussion ...... 147 5.6 Conclusions ...... 151 5.7 Acknowledgments ...... 152 5.8 References ...... 152 5.9 Figures ...... 159

CHAPTER 6: CELL- AND ANTIBODY-MEDIATED IMMUNITY INDUCED BY PATHOGEN-MIMICKING NANOVACCINES ...... 165 6.1 Abstract ...... 166 6.2 Introduction ...... 166 6.3 Materials and Methods ...... 168 6.3.1 Materials ...... 168 6.3.2 Monomer, polymer, and nanoparticle synthesis ...... 170 6.3.3 Nanoparticle functionalization ...... 171 6.3.4 Characterization of functionalized nanoparticles ...... 171 6.3.5 Animals ...... 172 6.3.6 Adoptive transfer and mouse immunizations ...... 172 6.3.7 T cell expansion and intracellular cytokine staining ...... 173 6.3.8 Antigenic challenge...... 173 6.3.9 Antibody titer ...... 174 6.3.10 Statistical analysis ...... 174 6.4 Results ...... 174 6.4.1 Nanoparticle characterization ...... 174 6.4.2 Carbohydrate-functionalized nanoparticles induced reduced antigen-specific CD8+ T cell expansion ...... 175 6.4.3 Carbohydrate-functionalized nanoparticles induced weak memory recall responses ...... 177 6.4.4 Nanovaccines induced robust antibody titers ...... 177 6.5 Discussion ...... 178 6.6 Conclusions ...... 184 6.7 Acknowledgements ...... 184 6.8 References ...... 184 6.9 Figures ...... 190 6.10 Tables ...... 196 CHAPTER 7: HIGH THROUGHPUT AUTOMATED SYNTHESIS OF PROTEIN- LOADED POLYANHYDRIDE NANOPARTICLE LIBRARIES ...... 197 7.1 Abstract ...... 198 7.2 Introduction ...... 199 7.2 Materials and Methods ...... 200 7.2.1 Materials ...... 200

7.2.2 Automated high throughput robot ...... 201 7.2.3 Polymer library synthesis ...... 202 7.2.4 Nanoparticle library synthesis ...... 203 7.2.5 Nanoparticle library functionalization ...... 204 7.2.6 Characterization of functionalized nanoparticle libraries ...... 205 7.2.7 Protein release kinetics...... 206 7.3 Results & Discussion ...... 206 7.3.1 Polymer and nanoparticle characterization ...... 206 7.3.2 Protein release from polyanhydride nanoparticles ...... 209 7.3.3 Nanoparticle surface functionalization and characterization ...... 211 7.4 Conclusion ...... 213 7.5 Acknowledgements ...... 213 7.6 References ...... 213 7.7. Figures...... 218 7.8 Tables ...... 222 CHAPTER 8: COMBINATION NANOVACCINES BASED ON HEMAGGLUTININ AND NUCLEOPROTEIN PROTECT AGAINST INFLUENZA CHALLENGE ...... 224 8.1 Abstract ...... 225 8.2 Introduction ...... 225 8.3. Materials and Methods ...... 227 8.2.1 Monomer, polymer, hydrogel and nanoparticle synthesis ...... 227 8.2.2 Animals ...... 229 8.2.3 Immunizations ...... 229 8.2.4 T cell analysis and flow cytometry ...... 230 8.2.5 Antibody titer ...... 232 8.2.6 Viral challenge ...... 232 8.2.7 Viral load quantification and clinical evaluation ...... 233 8.2.8 Statistical analysis ...... 234 8.3 Results ...... 234 8.3.1 Combination nanovaccine induced T cell expansion ...... 234 8.3.2 Nanovaccine formulations induced robust antibody titers ...... 236 8.3.3 Mice immunized with nanovaccine formulations had reduced viral titer upon challenge ...... 237 8.3.4 Immunized mice lost less weight following viral challenge ...... 237

vii

8.3.5 T cell expansion in the lung following viral challenge ...... 238 8.4 Discussion ...... 239 8.5 Conclusions ...... 244 8.6 Acknowledgements ...... 245 8.7 References ...... 245 8.8 Figures ...... 250 CHAPTER 9: CONCLUSIONS AND ONGOING/FUTURE WORK ...... 257 9.1 Conclusions ...... 257 9.2 Ongoing Research and Future Studies ...... 259 9.3 References ...... 263 9.4 Figures ...... 265

viii

ACKNOWLEDGEMENTS My graduate education was a personal achievement that is rooted in the contributions and support of others. First, it was with sincere gratitude that I would like to acknowledge my advisor and mentor Dr. Balaji Narasimhan. I am truly appreciative that I have been able to work in his laboratory under his guidance. His genuine optimism coupled with his scientific expertise has allowed me to grow into a more thoughtful and well-rounded scientist. Additionally, his strong interpersonal and communication skills are qualities that I have tried to emulate, and in doing so has greatly enhanced my professional development. I would also like to acknowledge my committee members Dr. Michael Wannemuehler, Dr. Marian Kohut, Dr. Ian Schneider, and Dr.

Jackie Shanks. They have provided me with great direction and sound technical support. A special thanks to Dr. Wannemuehler for intellectually stimulating discussions on Immunology and trying to get me to understand some of the fundamental mechanisms of biology.

Throughout the course of research, I have been able to cultivate strong friendships and great working relationships with Dr. Tim Brenza, Dr. Kathleen Ross, Dr. Shannon Haughney,

Dr. Julia Vela Ramirez, Adam Mullis, and Sean Kelly. The help and support from my undergraduate students: Lucas Dunshee and Akash Mitra are also acknowledged. Additionally, I would like to thank collaborators Dr. Paola Boggiatto, Danielle Wagner, Ross Darling, and

Matthew Jefferson. I wish nothing but success for incoming graduate students: Sujata Senapati and Benjamin Schlichtmann.

Most of all I would really like to thank my parents, Brian and Peggy Goodman, who constantly tried to instill the value of a good education. I would like to thank my siblings Nicole

Goodman, Lance Goodman, and Amber Goodman for their support. Many close friends made my time in Ames memorable including Brice Floyd, John Matthiesen, and Michael Nolte. ix

Although not always easy, my tenure at Iowa State was characterized by many enjoyable experiences which have continued to mold my personality and my outlook on life.

CHAPTER 1: INTRODUCTION It has been said that vaccination is one of the most effective medical discoveries ever (1).

Some estimates show that each year vaccination prevents 2.5 million deaths worldwide (2). The success of vaccines is most profoundly illustrated in a recent study, which showed dramatic decreases or complete elimination of common infectious diseases in the post-vaccine era that induced a great burden in the pre-vaccine era (Figure 1.1) (3). Some diseases have been completely eliminated, such as polio and smallpox, and others have seen >90% reduction in morbidity, including measles, hepatitis, and chickenpox.

The birth of vaccination began in the late 1700s when Edward Jenner used pustules from cows with cowpox to protect humans from the closely related smallpox (2). Cowpox induced disease in humans, albeit not nearly as severe as smallpox and more importantly protected humans from getting smallpox. Vaccines have evolved greatly since then. Later, it became common practice to inactivate microorganisms that caused a disease and inject them into humans to induce an effective immune response. Although these vaccines were effective, they exhibited high reactogenicity and inflammation at the injection site (4). Recent breakthroughs in vaccinology have allowed for development of a new generation of vaccines. These breakthroughs include viral propagation in cell culture, the use of recombinant expression systems to produce large quantities of antigen, and the utilization of novel adjuvants and delivery systems to amplify the immune response (5-7). These new generation vaccines are also enabling the delicate balance between safety and efficacy. In other words, the primary objective of any vaccine going to market today is to increase the immunogenicity without compromising safety.

A well-studied strategy to augment the immunogenicity of vaccines while maintaining safety is to encapsulate pathogen-specific antigens into biodegradable delivery vehicles. This

helps ensure that the antigen interacts with cells internalizing the delivery vehicle. In particular, synthetic biodegradable polymers, such as polyanhydrides, have shown to be promising candidates as both vaccine adjuvants and as vehicles to deliver antigen to immune cells (8-11).

This is because micro- or nanoparticles synthesized from these polymers serve as antigen carriers that stabilize their payload, while providing a controlled release for sustained antigen persistence

(12). Simultaneously they present molecular patterns that enhance the adjuvanticity and enable immunomodulation (13).

An alternative method to enhance the immunogenicity of vaccines is to actively target antigen presenting cells (APCs). Current vaccines on the market do not target cells (13).

However, many active targeting techniques under development rely on targeting pattern recognition receptors (PRRs), such as C-type lectin receptors or Toll-like receptors (14). PRRs are an important part of the innate immune systems that recognize conserved molecules from pathogens called pathogen-associated molecular patterns (PAMPs) (14). They are convenient to target because they are usually expressed on the surface or in the endosome of APCs and other immune cells, such as macrophages, B cells, and T cells. This makes them easily reachable with appropriately designed delivery systems (15; 16). Indeed, there have been numerous studies to determine the signaling pathways and outcomes of targeting PRRs (17; 18).

The overall goal of this thesis is to integrate active targeting of CLRs with the polyanhydride nanovaccine delivery platform to enable the design of targeted nanovaccines against respiratory pathogens, in particular, influenza A virus.

1.1 References 1. Jones KS. 2008. Biomaterials as vaccine adjuvants. Biotechnology progress 24:807-14

2. De Gregorio E, Rappuoli R. 2014. From empiricism to rational design: a personal perspective of the evolution of vaccine development. Nature reviews. Immunology 14:505-14

3. Herper M. 2013. How Vaccines Have Changed Our World in One Graphic.

4. Bachmann MF, Jennings GT. 2010. Vaccine delivery: a matter of size, geometry, kinetics and molecular patterns. Nature reviews. Immunology 10:787-96

5. Wong SS, Webby RJ. 2013. Traditional and new influenza vaccines. Clinical microbiology reviews 26:476-92

6. Ulmer JB, Valley U, Rappuoli R. 2006. Vaccine manufacturing: challenges and solutions. Nature biotechnology 24:1377-83

7. Wang K, Holtz KM, Anderson K, Chubet R, Mahmoud W, Cox MM. 2006. Expression and purification of an influenza hemagglutinin--one step closer to a recombinant protein-based influenza vaccine. Vaccine 24:2176-85

8. Kipper MJ, Shen E, Determan A, Narasimhan B. 2002. Design of an injectable system based on bioerodible polyanhydride microspheres for sustained drug delivery. Biomaterials 23:4405-12

9. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

10. Huntimer L, Wilson Welder JH, Ross K, Carrillo-Conde B, Pruisner L, et al. 2013. Single immunization with a suboptimal antigen dose encapsulated into polyanhydride microparticles promotes high titer and avid antibody responses. Journal of biomedical materials research. Part B, Applied biomaterials 101:91-8

11. Huntimer L, Ramer-Tait AE, Petersen LK, Ross KA, Walz KA, et al. 2013. Evaluation of biocompatibility and administration site reactogenicity of polyanhydride-particle-based platform for vaccine delivery. Advanced healthcare materials 2:369-78

12. Determan AS, Wilson JH, Kipper MJ, Wannemuehler MJ, Narasimhan B. 2006. Protein stability in the presence of polymer degradation products: consequences for controlled release formulations. Biomaterials 27:3312-20

13. Mallapragada SK, Narasimhan B. 2008. Immunomodulatory biomaterials. International journal of pharmaceutics 364:265-71

14. O'Neill LA, Bryant CE, Doyle SL. 2009. Therapeutic targeting of Toll-like receptors for infectious and inflammatory diseases and cancer. Pharmacological reviews 61:177-97

15. Bernasconi NL, Onai N, Lanzavecchia A. 2003. A role for Toll-like receptors in acquired immunity: up-regulation of TLR9 by BCR triggering in naive B cells and constitutive expression in memory B cells. Blood 101:4500-4

16. Muzio M, Bosisio D, Polentarutti N, D'Amico G, Stoppacciaro A, et al. 2000. Differential expression and regulation of toll-like receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells. Journal of immunology 164:5998-6004

17. Tacken PJ, de Vries IJ, Torensma R, Figdor CG. 2007. Dendritic-cell immunotherapy: from ex vivo loading to in vivo targeting. Nature reviews. Immunology 7:790-802

18. Geijtenbeek TB, Gringhuis SI. 2009. Signalling through C-type lectin receptors: shaping immune responses. Nature reviews. Immunology 9:465-79

1.2 Figures

Figure 1.1 Vaccines have contributed to a dramatic reduction of some of society’s worst diseases. Reprinted from (3).

CHAPTER 2: LITERATURE REVIEW

This chapter summarizes the major advancements made in the fields of polymeric vaccine delivery systems (Section 2.1) and targeted delivery systems (Section 2.2). This review summarizes the wide array of polymers that have been used as delivery vehicles, including polyesters, polyanhydrides, natural based polymers, and liposomes. A number of techniques have been used to actively target drugs and vaccines to cells or tissues, and these are comprehensively reviewed herein. The chapter concludes with a discussion on the influenza virus and the design of novel vaccines against influenza (Section 2.3). New generations of influenza vaccines are on the horizon and they often utilize delivery systems as well as active targeting mechanisms to provide enhanced protection from the virus.

2.1 Polymeric Vaccine Delivery Systems

2.1.1 Introduction Biodegradable micro- and nanoparticles for vaccine delivery can be synthesized from a number of building blocks including proteins, polysaccharides, or synthetic polymers (1). They can function effectively as vehicles to deliver antigen and their unique properties can also impart adjuvant-like characteristics that enhance the resulting immune response.

The in vivo performance of nanomedicines is a function of their hydrophobicity, surface modification, surface charge, and particle size (2). Hydrophobicity has two important roles. It influences the rate of payload release from the nanoparticles and it affects the in vivo circulation time of the nanoparticles. Biodegradable delivery vehicles that are hydrophobic inhibit water from penetrating into the polymer matrix, thus slowing down degradation and payload release. It has also been suggested that hydrophobicity is a “danger signal” to immune cells (3). However, hydrophobic particles are known to be rapidly opsonized and cleared by mononuclear phagocytic 7

cells (2). As a result, particles used for drug delivery requiring long circulation times generally have their surfaces conjugated with a more hydrophilic moiety, such as polyethylene glycol (4).

Surface modification can be used to design particles to target a specific receptor on a cell, while surface charge can help particles interact with oppositely charge cellular membranes. It is well known that a positive surface charge is preferable as it is often shown to increase nanoparticle internalization (5). Nanoparticles with sizes below 1000nm have important properties for drug and vaccine delivery. They can penetrate capillaries and tissue gaps to reach their target organs.

In particular, particles between 20-100nm have been shown to be optimal for entering the lymphatic system (6). They are subcellular in size, can be readily internalized, and still provide controlled release of drugs or antigen that ensure stability and improved utility (7).

Drugs and antigens are either bound to the surface of delivery vehicles or encapsulated inside the vehicle (2). Both non-covalent adsorption and covalent attachment of drugs to the surface of vehicles have been reported (8-10). These systems have been shown to present particulate antigen to B cells, thus amplifying the humoral immune response. However, this approach has important limitations when compared to delivery vehicles that encapsulate antigens. First, the antigen is subjected to environmental conditions, which makes it susceptible to degradation. Second, the delivery vehicle releases its payload all at once, whereas encapsulated antigens tend to release their payload more gradually as the delivery vehicle degrades. A sustained release of antigen has been shown to be optimal to prime adaptive immune responses (11; 12).

There are a number of techniques to characterize nano-based delivery vehicles. These include imaging techniques, such as scanning or transmission electron microscopy, which characterize the morphology and size of the delivery vehicle (13). Zeta potential experiments can

be used to measure the average surface charge of the delivery vehicle by analyzing the mobility of charged particles in response to an electric potential (14). Particle size and size distribution can be determined by dynamic light scattering, which analyzes the light scattered by the

Brownian motion of the particles (14).

Current research has shown that nanomedicines have the potential to serve as an excellent platform for the development of next generation vaccines (15; 16). In particular, several classes of nanomedicines derived from polyesters, polyanhydrides, natural polymers, and liposomes, are already in various stages of preclinical and clinical trials for both vaccine and drug delivery.

These different polymeric chemistries are reviewed next.

2.1.2 Polyesters

Biodegradable polyesters first rose to prominence for biomedical applications in the early

1970’s when researchers successfully demonstrated that these polymers were ideal for sutures and fibers (17). These aliphatic polyesters are thermoplastic and can be easily formed into a number of shapes. Early research demonstrated that polyesters exhibited great biocompatibility, low toxicity, and had predictable degradation rates leading to their approval by the U.S. Food and Drug Administration (FDA) for therapeutic uses in humans (18). These beneficial properties have been exploited since to make polyesters among the most attractive biomaterials for drug and vaccine delivery.

Biodegradable polyesters are composed of lactic acid or glycolic acid monomers and copolymers thereof (i.e., poly(lactic-co-glycolic acid) or PLGA) (Table 2.1). Poly(lactic acid)

(PLA) has a chiral α-carbon that in stereochemical terms is described as D or L and the enantiomer type can greatly affect the properties of the polymer.(19). Poly L-lactic acid is more crystalline while poly D-lactic acid has disordered polymer chains and so it is amorphous (19).

Poly(glycolic acid) (PGA) has no methyl groups and is also crystalline (Table 2.1). PLGA usually has an equal mixture of D- and L- lactic acid enantiomers (19). The polymers are synthesized using a ring-opening melt condensation reaction of lactide and glycolide cyclic dimers. Polymerization proceeds for 2-6 hours at 175°C in the presence of organotin catalysts

(17). It is known that environmental humidity and the amount of free acid impurities in the melt significantly affect polymer synthesis (17). Both need to be at low levels to achieve high molecular weight polymers. Today, PLA, PGA, and various PLGA copolymers can be purchased commercially with a variety of molecular weights.

In aqueous solutions, polyesters degrade by the hydrolysis of their ester linkages (Figure

2.1) (20). This forms monomeric lactic acid or glycolic acid that can be safely metabolized through the Krebs cycle. Crystallinity and water uptake are important factors that influence the degradation rate. The methyl groups on PLA make it more hydrophobic than PGA, meaning they absorb less water (19). As a result PLA homopolymers have a slower degradation rate than

PLGA or PGA. Generally increasing the GA content in PLGA increases the degradation rate of the copolymer. The one exception is 50:50 PLGA, which degrades the fastest. Polyesters also degrade more slowly in basic solutions compared to acidic solutions because their degradation is acid-catalyzed (21). Polyesters are also known to undergo bulk erosion, in which the rate of hydration of water molecules diffusing into the polymer matrix is faster than polymer hydrolysis or solubilization (Figure 2.2) (20). This bulk erosion behavior makes PLGA erosion more unpredictable than surface eroding polymers.

Polyesters are soluble in several solvents, including acetone, chlorinated solvents, and tetrahydrofuran, resulting in a number of methods to synthesize nano- or microparticles (19; 22).

The most common technique to prepare polyester particles is the single emulsion solvent

evaporation technique (23). Polymer is dissolved in a volatile organic solvent while the drug or antigen usually has to be dispersed because it does not dissolve in organic solvents. This solution is emulsified in a large volume of water with the presence of a surfactant, such as polyvinyl alcohol or polysorbate-80 (13). The organic solvent is then evaporated inducing particle formation, upon which the particles are collected through centrifugation. Water-soluble drugs or antigens that don’t disperse well in organic solvents can be encapsulated into polyester nanoparticles through a double emulsion process (13). Here the polymer is again dissolved in the organic solvent; however, the antigen is dissolved in water and the two are mixed using rapid stirring to create an emulsion. This solution is then added to a large volume of water that contains a surfactant and is continuously emulsified. Evaporating the organic solvent results in nanoparticles.

PLGA nanoparticles are efficiently internalized by phagocytic cells and deliver their payload to the lysosomal compartments of cells, through pathways independent of clathrin- coated endocytosis (24; 25). Some researchers have reported that PLGA nanoparticles were efficient at escaping the endosome and making it to the cytoplasm (26). Similarly another study showed that ovalbumin (OVA) loaded PLGA particles stimulated dendritic cells to produce IL-2 and led to sustained MHC presentation as compared to soluble antigen (27). This could potentially lead to better T cell responses in vaccine development. PLGA delivery systems have the capability to co-encapsulate antigen with other adjuvants so that both are internalized by the same cells (28). Polyesters have also been used to encapsulate nucleic acids and protect them from nucleases (29). However, due to their negative charge, nucleic acids do not disperse readily in the organic solvents during nanoparticle synthesis and so they are often complexed with cationic polymers to improve encapsulation efficiency. In one such study, researchers complexed

a cationic polymer with siRNA to enhance the encapsulation efficiency from 43% to 79% (30).

Lastly, in addition to human vaccines, PLGA nanoparticles have been explored as oral delivery devices for fish (31).

Perhaps the most significant detriment to PLGA-based delivery devices is that when they degrade, their acidic degradation products can be trapped in the polymer matrix, reducing the local pH of the medium to as low as 3 (32). This can result in hydrolysis of the encapsulated payload, particularly proteins (32; 33). Degraded proteins in vivo can have detrimental effects, including loss of activity or function. The acidic microenvironment has been alleviated by incorporating basic inorganic salts into the particles, such as magnesium hydroxide or sodium bicarbonate (32; 33). PLGA particles also tend to exhibit a burst release in the early stage of incubation with aqueous solutions. This is believed to be due to surface adsorbed or poorly encapsulated protein.

Poly(ε-caprolactone) (PCL) is another aliphatic polyester that has been investigated for drug delivery applications (Table 2.1) (34). Unlike PLA or PGA, is has a very low glass transition temperature of about -60°C and a low melting point of 63°C. This makes melt encapsulation a feasible technique to form biomedical devices encapsulating proteins (35). In contrast, PLA, PGA, and their copolymers have glass transition temperatures that are above room temperature and melting points above 100°C (18). PCL degrades very slowly in water and is used for long-term drug delivery usually in excess of one year (36). This slow degradation can be advantageous as PCL degradation does not generate a locally acidic environment unlike

PLGA degradation. If faster degradation rates are needed, PCL is sometimes formulated as a blend with PLGA and then synthesized to form nanoparticles (36).

2.1.3 Polyanhydrides

The study of polyanhydrides as a platform for drug delivery originated in the early 1980’s by Langer and co-workers (37). The presence of a hydrophobic backbone together with water- labile anhydride linkages enables great control over degradation kinetics and subsequently drug release. This has stimulated over three decades of research on the use of these surface-erodible polymers for drug and vaccine delivery.

An anhydride polymer is composed of multiple monomer dicarboxylic acids (Figure 2.1)

(38). The physical properties of each monomer are unique and they impart their characteristics on the resulting polymer. Aliphatic diacids and aromatic diacids are commonly used monomers to synthesize polyanhydrides. Aliphatic polyanhydrides consist of a straight or branched chain of saturated carbon atoms that make up the backbone of the polymer (Table 2.1). These homopolymers are highly crystalline, upwards of 60% for poly(sebacic anhydride) (PSA), and have melting points ranging from 50-90°C (39; 40). Polyanhydrides are generally eliminated from the body over a period of several weeks after in vivo injection (41). Their degradation products, which are dicarboxylic acids, are non-toxic and easily metabolized by the body (Figure

2.1). Polymerization with aromatic diacids results in melting points greater than 100°C, insolubility in most organic solvents, and slow degradation rates that can last over a year due to the increased hydrophobicity associated with the aromatic rings (38). Adding methyl groups to the backbone of an aromatic diacid decreases the melting point of the corresponding homopolymers (41). Aromatic-based anhydrides in vivo have also proven to be safe and non- toxic; however, their degradation products are not metabolized and are excreted instead (41).

There are numerous other families of diacids, including branched diacids, unsaturated diacids and fatty acid-derived diacids, but they are less commonly utilized in drug delivery.

Synthesizing anhydride polymers is most commonly performed through melt polycondensation. Melt polycondensation generates high molecular weight polymers (up to 100 kDa). The dicarboxylic acids are acetylated by refluxing in excess acetic anhydride and melted at

180°C for 2-3 hours with a high vacuum to remove the condensation product and push the polymerization forward (38). Other synthesis routes include dehydrochlorination, dehydrative coupling, and ring opening polymerization (38). The physical properties of these polymers can be further tailored by copolymerizing different types of monomer diacids (42). Typically, aliphatic and aromatic diacids are copolymerized together to adjust drug release times or hydrophobicity.

When exposed to water, polyanhydrides undergo hydrolysis of their anhydride bonds

(Figure 2.1). One of the advantages of polyanhydrides is that they are surface-eroding polymers, enabling more predictable release kinetics (Figure 2.2). This occurs because the polymeric chains at the interface of the polymer and water break down and release the drug before water can penetrate into the polymeric matrix (40; 43). This degradation is controlled by a number of factors. For example, aromatic homopolymers degrade several orders of magnitude slower than aliphatic homopolymers (41; 44). As a dry powder, these aromatic polymers have been able to retain their original molecular weight for over a year, while aliphatic polymers have exhibited significant degradation during this time (37). Generally, in aqueous solutions hydrophobic polymers degrade more slowly than less hydrophobic polymers because there is reduced water permeability into the polymeric matrix (43). Additionally, semicrystalline polymers degrade more slowly than amorphous polymers. The pH of the solution can greatly affect polymer degradation. Polyanhydride degradation is base-catalyzed and their degradation rates increase with higher pH. Copolymer degradation is more complex. In a random A-B copolymer there are

three bond types, A-A, B-B, and A-B, and the degradation rates of all three bonds need not be equal. This leads to the formation of microphases and non-homogeneous degradation (38).

Degradation products resulting in the formation of dicarboxylic acids lowers the local pH.

However, unlike polyester degradation, this decrease in pH is milder (45).

Polyanhydrides have been utilized as implantable and injectable vehicles for drug delivery. Implantable systems consist of a wafer made by mixing the drug with the polymer powder and forming the wafer through compression molding or melt compression. The most well-known polyanhydride system, the Gliadel® wafer, is a compression-molded wafer (46). The

Gliadel® wafer is an FDA-approved implant, consisting of a copolymer of sebacic acid (SA) and

1,3-bis(p-carboxyphenoxy)propane (CPP). It is post-surgically implanted into the brain of patients with malignant glioma, where it degrades and releases a chemotherapeutic, carmustine, to inhibit tumor growth (47).

Nanoparticles and microparticles based on polyanhydrides have been synthesized using a number of techniques including spray drying, solvent extraction, or cryogenic atomization

(Figure 2.3) (48; 49). Solvent extraction can involve the use of a single emulsion or a double emulsion. In single emulsion systems, such as the solid/oil/oil (s/o/o) method, the drug and polymer are dissolved in a solvent and then dispersed in a non-solvent, precipitating the micro- or nanoparticles (Figure 2.3) (40; 49). Other nanoprecipitation techniques involve double emulsions, such as the water/oil/water (w/o/w), and water/oil/oil (w/o/o) emulsions. These are used when a drug is poorly soluble in the oil solvent. The drug is first dissolved in water, while the polymer is still dissolved in a solvent. The two are emulsified through vigorous mixing and the particles are precipitated after exposure to the non-solvent, which is either water or oil. The s/o/o method is convenient because it provides high encapsulation efficiencies and avoids a

water/oil interface that can cause protein migration and unfolding (48; 49). Spray drying involves using a volatile solvent to dissolve the polymer and drug (Figure 2.3). The solution is atomized through a spray dryer with a stream of hot air. This induces the solvent to evaporate leaving behind precipitated particles. An alternative method to make particles is through cryogenic atomization. In this method, the polymer and drug are dissolved in a solvent and pumped through an ultrasonic nozzle into a bath of liquid nitrogen that is on top of frozen ethanol (49). The ethanol slowly thaws and as the frozen particles fall into it, the solvent diffuses into the ethanol, inducing precipitation of the particles.

Polyanhydride micro- and nanoparticles enhance protein stability and have been shown to release biologically and functionally active protein antigens that maintain their primary, secondary, and tertiary structure (45; 50-53). Recent studies from our laboratory demonstrated that the pneumococcal surface protein A, the hemagglutinin protein from the H5 influenza virus, and gp41 from the HIV virus were stably released from polyanhydride nanoparticles, demonstrating their value as a viable delivery platform for vaccines (54-56). These studies also demonstrated that the chemistry of polyanhydrides can be tailored to provide a sustained release of antigen. Vaccine formulations based on such nanoparticles allow for providing prolonged exposure of antigen without needing more administrations.

In addition to stabilizing protein antigens, polyanhydride delivery vehicles have shown the ability to adjuvant the immune response to antigen (57). Several studies have shown enhancement of both cellular and humoral immune responses in a single dose, immunomodulation, and dose sparing using polyanhydride nanovaccines (53; 58-62).

Additionally, the internalization by innate immune cells of hydrophobic polyanhydride formulations was enhanced because the surface patterns of the hydrophobic chemistries

mimicked PAMPs and triggered danger signals to APCs (63). The majority of these particles were localized in the endosomal regions of the cell, which is an important pathway for antigen loading on to major histocompatibility complex II molecules.

2.1.4 Natural polymers and polysaccharides

Natural polymers have been extensively studied as candidates for enhancing the delivery of drugs and vaccines. They are derived from plants, animals, and microbes (2). Typically they are biodegradable and have higher biocompatibility than synthetic polymers (7). Two of the most prominent natural polymers used to make nanoparticles or microparticles are chitosan and gelatin.

Chitosan is a linear polysaccharide with long polymer chains (50-200 kDa) (Table 2.1)

(64). It is produced commercially by the deacetylation of chitin, which is a natural biomaterial from the shells of crustaceans and insects. This polymer is composed of β-(1-4) linkages of glucosamine and N-acetyl glucosamine and has nitrogen groups on its polymer backbone making it cationic (64). These nitrogen moieties impart important properties to chitosan that have contributed to its success in drug delivery. Chitosan is non-toxic, stable, easily sterilized, and biodegradable, making it a strong candidate for the controlled release of vaccines (65).

Chitosan is water soluble only at low pH, but it can be methylated to form a derivative called N-trimethyl chitosan (TMC), which has a positive charge regardless of pH and is more stable in aqueous solutions (66). Particles are most commonly synthesized through coacervation, ionic gelation, or spray drying. The coacervation technique takes advantage of chitosan being insoluble in alkaline solutions. Here the dissolved solution is passed through a compressed air nozzle to form droplets and precipitated into an alkaline solution (67; 68). Following purification through centrifugation or filtration, the particles can be hardened using a crosslinking reagent,

such as gluteraldehyde, to control release rate. Spray drying chitosan particles is very similar to that of polyanhydrides except that the polymer is dissolved in acetic acid (Figure 2.3) (67). The solution is atomized in hot air leading to the formation of small droplets that evaporate and inducing particle formation. In ionic gelation, the chitosan and antigen are dissolved in acidic solution and it is added dropwise to a polyanionic solution (66). The two oppositely charged chemicals complex and precipitate. Due to a lack of covalent bonds, these particles usually have poor mechanical strength and rapidly release their encapsulated payload (67).

A major advantage of chitosan-based nanoparticles is that they are mucoadhesive, making them an excellent candidate for intranasal immunizations (68). Mucins in the intranasal mucosal tissue are negatively charged and the strong electrostatic interactions between mucins and the positively charged chitosan particles help increase the retention time of the particles and encapsulated antigen (69). It has also been suggested that chitosan has the capacity to loosen the tight junctions formed by epithelial cells, allowing the particles to more easily cross epithelial barriers making for enhanced drug delivery (67). Chitosan particles and chitosan-coated particles administered intranasally to mice have induced both systemic and local immune responses (70;

71). For example, one study compared the immune responses of antigen encapsulated in PLGA nanoparticles coated in chitosan or chitosan nanoparticles to that of soluble antigen upon intranasal administration to mice (66). The antigen released from the chitosan nanoparticles had the longest retention time as measured by the in vivo fluorescence intensity of dye-conjugated antigen and also generated the highest serum antibody titers and IgA titers. Other studies have also used chitosan-coated particles for nasal administration. PCL particles were coated in chitosan and delivered intranasally to mice in order to induce an immune response against

influenza (72). One shortcoming of chitosan is that its molecular weight is extremely variable and hard to control.

Another important natural polymer used in drug delivery is gelatin. Gelatin is a polyampholyte and has both positively charged and negatively charged functional groups (73). It also has hydrophilic components that lead to swelling when these particles are incubated in aqueous solutions, enabling easier diffusion and release of antigen. However, the swelling and release rate can be controlled by crosslinking with glutaraldehyde (74). In a study using microspheres, gelatin particles with 0.06% glutaraldehyde released 64% of its payload in 30 minutes. When the amount of glutaraldehyde was increased to 0.18%, the release rate was 37% of the payload in the same time (75). Even so, these are relatively high release rates, which is a limitation in using gelatin based drug formulations for drug and vaccine delivery applications.

Gelatin is a protein-based polymer derived from the hydrolysis of collagen from either bovine or swine origin (76). The hydrolysis is either base-catalyzed or acid-catalyzed, which in turn affects the functional groups that are present. The diversity of functional groups available makes gelatin a good candidate for surface functionalization (77). It also has low antigenicity, is biodegradable, and has been well studied for drug delivery (78; 79). In addition to diffusion and erosion, gelatin particles can be broken down by proteases (78). One study showed that only about 10% of doxorubicin loaded into gelatin nanoparticles were released in 72 hours when incubated in saline solution, but upon adding proteases this release rate was immediately increased several-fold (80).

2.1.5 Liposomes

Liposome-based vehicles have been clinically approved for drug delivery. Astellas

Pharmaceutical Corporation has designed a liposome-based formulation to increase the solubility

and targeting of the anti-fungal drug amphotericin B (81). Other manufacturers use liposomes to limit the non-specific toxicity of anti-cancer drugs such as doxorubicin and daunorubicin (81). In terms of vaccine delivery, liposomes are now being utilized in clinical trials. One example is the cationic liposome, CAF01, which is used as a delivery vehicle with a tuberculosis (TB) antigen in a TB vaccine (82; 83). This liposome is formed with a synthetic cationic lipid along with a synthetic lipid derived from the cell wall of Mycobacterium tuberculosis to adjuvant the immune response. This vaccine induced a strong Th1-mediated immune response (82; 83).

Liposomes are bilayered vesicles that form spontaneously when lipids are incubated in aqueous solutions (Figure 2.4) (84). Their release kinetics as well as their ability to adsorb or encapsulate antigen depend greatly on their composition, charge, and size. Liposome formation is most affected by the geometry of the individual lipid monomers. Typically, the lipid monomer is composed of a large hydrophilic head group followed by long double hydrocarbon chains (85).

Hydrocarbon chains that are saturated and long exhibit more van der Waals forces with neighboring chains leading to more rigid liposomes. Unsaturated or short hydrocarbon chains give the liposome more fluid like properties. Liposome fluidity can also be altered by incorporating sterols, such as cholesterol, into the liposome membrane (86). This fluidity can affect the immune response. Investigators have found that more rigid liposomes can form long term depots and do not disperse as readily from the site of injection as more fluid liposomes, resulting in more robust immune responses (81).

Liposomes are most commonly made with the neutral lipid monomer, phosphatidylcholine, while incorporating smaller amounts of cationic or anionic lipid monomers in the membrane bilayer to get a charged delivery vehicle (86). Several studies have shown that cationic liposomes are more immunostimulatory than anionic liposomes (87; 88). All liposomes

have an aqueous core that allows encapsulation of drugs, proteins, or DNA (Figure 2.4). A disadvantage of liposomes is their stability. In vivo, their release times have been reported to range from minutes to hours depending on the composition (85). Methods have been developed to obtain lyophilized liposomes to prevent their degradation and release of antigen in aqueous solutions during storage (85). This necessitates the use of excipients, including cryoprotective agents, such as trehalose. Additionally upon in vivo administration, liposomes have been shown to be rapidly consumed by the mononuclear phagocyte system (MPS) because studies have shown that nearly 80% of the dose is adsorbed by the MPS within 30 minutes of intravenous administration (85).

2.2 Targeted Delivery of Vaccines*

2.2.1 Introduction

Vaccines that can direct their action to a specific target on antigen presenting cells

(APCs) can overcome the lack of immunogenicity associated with subunit vaccines. This

“rational design” approach involves targeting specific receptors that have the ability to trigger immune activating and signaling events or to induce enhanced cellular internalization and processing of antigen. Pattern recognition receptors (PRRs) make excellent targets in the rational design of vaccines as they play an integral role in the innate host defense by recognizing conserved pathogen associated molecular patterns (PAMPs). Targeted vaccines can provide several advantages, including cost effectiveness, dose sparing, immune modulation, and less toxicity than alternative systems. Dose sparing and cost effectiveness correlate with each other in that vaccines that target PRRs have the capability to induce robust immune responses using less antigen than conventional vaccines. Rationally designed antigens can be targeted to specific

*Accepted for publication in Annual Review of Biomedical Engineering (to appear in March 2016)

receptors or tailored to induce specific cytokine profiles that modulate the immune response (89).

Vaccines that target APCs cells while avoiding a systemic response could also reduce toxicity to the host.

2.2.2 Types of PRRs

2.2.2.1 C-Type Lectin Receptors (CLRs) Most CLRs are membrane bound proteins that contain one or more carbohydrate recognition domains (CRDs) for binding to carbohydrates. With over 60 types of receptors,

CLRs represent a diverse receptor family that have been shown to mediate host biological events and contribute to the innate immune response (90). CLRs are often associated with a Ca2+ dependent mechanism of binding; however, some CLRs such as Dectin-1 have been shown to bind its carbohydrate ligand independent of Ca2+.

The macrophage mannose receptor (MMR) has eight CRDs located sequentially on one polypeptide chain and it is most efficient at recognizing single or low order mannose moieties that are present on yeast, fungi, viruses, and many bacteria including Mycobacterium tuberculosis and Streptococcus pneumoniae (Figure 2.5) (91). The MMR is most commonly expressed on immature DCs, tissue macrophages, and some epithelial cells, whereas circulating monocytes have very low levels of expression (Table 2.2) (92). It acts as an endocytic receptor that internalizes its ligand through coated pits. One specific example of its role in the immune response is the induction of maturation in human peripheral blood mononuclear cells, leading to a strong Th17 adaptive response, by mannan derived from Candida albicans (93). By blocking

MMR expression on DCs with siRNA, T cell activation was inhibited, preventing CD4+ cells from producing IL-17 and IFN-γ. There is increasing evidence to support that the MMR has a

“dual role” as an important PRR for the innate immune system and in homeostasis by the

clearance of endogenous ligands. For example, it has been shown that the blood of MMR- deficient mice showed delayed clearance of lysosomal enzymes that contained mannose side chains (94).

A soluble C-type lectin that is closely related to the MMR is the mannose binding lectin

(MBL). Each single polypeptide chain has one CRD, but usually three chains come together to form a collagen helix resulting in a single structure with three CRDs at its tip. These structures can further form multimers containing different numbers of structural subunits (95). Like the

MMR, the MBL has specificity for mannose, N-acetylglucosamine (GlcNAc), and fucose. Each

CRD binds with relatively low affinity and so ultimately it is the clustering of multiple CRDs that gives the MBL a strong avidity. The MBL, along with other soluble lectins, such as ficolin, have been shown to bind to bacterial peptidoglycan and fungal derived β-glucans (96). This induces the binding of MBL-associated serine proteases (MASPs) to promote the activation of the lectin pathway of the complement system (96; 97).

Similarly to the MMR, DEC-205 is another CLR used for targeting. It does not have a signaling motif on its cytoplasmic tail; however, there are two separate motifs that are believed to influence endocytosis and intracellular trafficking (Figure 2.5) (98). This includes a tyrosine- based motif as well as a triacidic cluster in contrast to the MMR that only has the tyrosine-based motif. It is believed that this triacidic cluster explains why DEC-205 transports its ligand to late endosomes whereas the MMR trafficking is mostly restricted to early endosomes (99). As a result targeting through DEC-205 is more efficient at presenting antigen than through the MMR

(99). In humans, DEC-205 is also readily expressed on DCs, particularly on myeloid DCs, moderate levels in B cells and low levels on T cells as well as natural killer cells (100). Unlike the MMR where DC expression is down-regulated upon maturation, DEC-205 expression is

generally up-regulated, which seems to imply that it has an important function in mature DCs

(Table 2.2) (101). Ligands taken up by DEC-205 mediated endocytosis have been shown not only to present antigen on MHC class II, but also on MHC class I, including cross presentation of antigen (102).

The dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-

SIGN) is a CLR expressed on immature DCs in peripheral tissues, such as the skin, gut, and lung, and its expression is down-regulated upon maturation (Table 2.2) (103). As a PRR it is mostly known for recognizing high mannose structures on surface glycoproteins of bacteria, fungi, and viruses including gp120 on HIV and glycoprotein E on the Dengue virus (104). DC-

SIGN has one CRD that is used for antigen recognition, cell transport across epithelial barriers and for interacting with other cells such as neutrophils. Studies have shown that the single polypeptide chains of DC-SIGN form tetramers with each other thus allowing each structure to present four CRDs (Figure 2.5) (105). DC-SIGN has a cytoplasmic tail consisting of three motifs, which include the tyrosine-based motif, a triacidic cluster, and a dileucine motif. Azad and co-workers inserted point mutations into each of the motifs individually to determine their function (106). They found that mutating the triacidic cluster reduced the surface expression of

DC-SIGN most significantly and it also had a 90% reduction in the phagocytosis of mannosylated-coated beads. Cells with the mutated dileucine motif had a small defect in phagocytosis and tended to retain antigen on its surface suggesting a role in internalization (106).

Mutations to the tyrosine motif had little effect on DC-SIGN expression or function. The overall conclusion was that these triacidic motifs are essential for DC-SIGN surface expression and that they play an important role in the uptake of antigen to the late endosome for fusion with the lysosome.

Dectin-1 is a proven phagocytic PRR for fungal β-glucans and it is expressed predominantly on DCs, macrophages, and neutrophils (Table 2.2). In particular, it is expressed on splenic macrophages, Kupffer cells, and macrophages in the lamina propria of the gut villi

(107). The cytoplasmic tail of Dectin-1 is unique to CLRs in that it has an immunoreceptor tyrosine-based activation motif (ITAM) that allows it to transduce signals into the cell and mediate internalization (108). Researchers have shown that Dectin-1 acts as a powerful PRR that can induce the innate immune system through Syk kinase signaling (Figure 2.5). This results in the production of reactive oxygen species and inflammasome formation. Additionally, it has been shown that Dectin-1 influences the adaptive immune response by inducing the secretion of

IL-1β and IL-23p19, which polarizes the T cell response to that of a TH17-mediated response

(109).

2.2.2.2 Toll-like Receptor (TLRs) TLRs are essential PRRs of the innate immune system and while they are not as diverse as their CLR counterparts, they remain a highly conserved class of receptors across multiple species. A total of 10 TLRs have been identified in humans and 12 in mice (110). Their main function is in signal transduction to induce the release of inflammatory cytokines or Type 1 interferons that can stimulate the adaptive response. TLRs are transmembrane proteins expressed on the surface of the plasma membrane or within endosomes. Many TLRs that recognize

PAMPS are located on the plasma membrane, while those that recognize nucleic acids are expressed on the surface of endosomes (111). Intracellularly, each TLR has a single

Toll/interleukin-1 receptor (TIR) domain similar to that of the IL-1 receptor that allows for downstream signaling. The extracellular portion contains leucine-rich repeats shaped like a horseshoe (Figure 2.5). Monocytes and myeloid DCs express TLR2 and TLR4 while plasmacytoid DCs are known for expressing TLR7 and TLR9 in higher abundance. They are

expressed at very low levels in B cells; however it has been found that TLR9 and TLR10 are up- regulated following triggering of the B cell receptor (112). Additionally TLRs are expressed on memory B cells (112). There is evidence showing that TLR2 is highly expressed on activated human T cells and memory T cells, suggesting a role as a co-stimulatory receptor of T cells and a role in the maintenance of memory T cells (113). At the same time, activated and memory T cells failed to respond to TLR4, TLR3, or TLR9 ligands.

The most notorious TLR is TLR4, because it is triggered by gram negative bacterial lipopolysaccharide (LPS). LPS induces a pro-inflammatory response through a receptor complex composed of MD-2 and TLR4, ultimately triggering the MyD88 dependent pathway and the activation of NFκB or the TRAM adaptor and activating interferon response factor 3 (IRF3)

(Figure 2.5) (114). There are numerous studies citing the importance of TLR4 activation in generating a protective immune response against pathogens (115-117). At the same time overstimulation of TLR4 can have adverse consequences and potentially lead to sepsis. In particular, fever and seizures have been reported with the pertussis whole cell vaccine because it triggers TLR4 (116). Other TLR4 agonists include the vaccine adjuvant, monophosphoryl lipid

A, and a number of heat shock proteins (118; 119).

TLR2 is important in the host response against gram positive bacteria and yeast because it binds a wide array of peptidoglycans, zymosan, and lipopeptides. It is unique in that it doesn’t form homodimers with itself, but instead forms heteromers with either TLR1 or TLR6 in order to initiate a signaling cascade (120). TLR2 signals through the MyD88 adaptor protein only when the Mal and MyD88 adaptor proteins bind to the TIR domains on the heterodimer (Figure 2.5).

Flagellin, originating from the filaments on the flagella of proteobacteria, binds to each

TLR5 and triggers the assembly of a homodimer allowing for MyD88-dependent signaling through the TIR domains (Figure 2.5) (121). Specifically TLR5 binds to conserved regions on the flagellin that are essential to the bacteria’s ability to form a filament. This suggests that the

TLR is well adapted to withstand mutations from the flagellin (122).

Unlike TLR4, TLR2 or TLR5, TLR3 and TLR 9 are endosomal TLRs (Figure 2.5). TLR3 recognizes double stranded (ds) RNA and thus plays an important role in the immune response against viruses, while TLR9 responds to unmethylated DNA, which is generally bacterial DNA

(Table 2.2). Poly(I:C) is typically the standard dsRNA immunostimulant that interacts with

TLR3, but it has been demonstrated to be slightly toxic and so modified versions that have reduced toxicity, such as, poly(ICLC) are often used (123). It has been determined that this receptor interacts with the sugar and phosphate backbones rather than the nucleotides of the dsRNA, which is why it is not specific for sequences (124; 125). The focus of current research is to utilize TLR3 and TLR9 ligands to induce more balanced immune responses in order to reduce

Th2-dependent pathways (126).

2.2.2.3 RIG-I Like Receptors (RLRs) There are three types of RLRs which include RIG-I, MDA5, and LGP2 and all are cytosolic PRRs (127). RIG-I and MDA5 have been shown to recognize RNA from different viruses and can initiate important host defense mechanisms against RNA viruses (128). This contrasts with LGP2, which has no signaling domains, but has been shown to be both a positive and negative regulator of RIG-I and MDA5 (129).

RIG-I has two caspase activation recruitment domains (CARD) which are the signaling domains of the receptor, and they activate the transcription factors NF-κB, AP-1, and IRF3

(Figure 2.5) (127). These act in concert to induce production of Type 1 interferons. RIG-I predominantly recognizes short ssRNA and dsRNA sequences if they contain a triphosphate at the 5’ end of the molecule and binds to longer dsRNA molecules up to several hundred base pairs long (130). Additionally RIG-I is indirectly the receptor for the adjuvant poly(dA:dT) when the DNA sequence is converted to RNA with a 5’-triphosphate through RNA polymerase III

(131). Initially RIG-I has a closed conformational structure where the two CARD domains are folded along the helicase. This prevents the CARD domains from accidently triggering downstream signaling. A conformational change occurs upon dsRNA binding to the CTD and helicase which is dependent on ATP hydrolysis (127; 132). MDA5 has a similar structure and signaling profile as RIG-I, but it primarily binds to long dsRNA molecules, including Poly(I:C)

(133).

2.2.2.4 NOD Like Receptors (NLRs) The NOD family is a diverse set of cytosolic receptors that specialize in recognizing

PAMPs against intracellular bacterial pathogens. They are able to bind to flagellin, peptidoglycans, or bacterial toxins, while others detect intracellular potassium levels to determine if there was a loss of host membrane activity (Table 2.2) (134). There are 23 NLRs in humans and 34 in mice and they are mostly expressed on macrophages and neutrophils (135).

There is also significant expression on epithelial cells and on specialized cells, such as Paneth cells in the gut, where they have a critical role in maintaining homeostasis with commensal bacteria (136).

Two predominant NLRs are NOD1 and NOD2. Both bind to fragments of peptidoglycan as NOD1 has specificity for meso-diaminopemilic acid (DAP) while NOD2 binds to muramyl dipeptide (MDP) (Table 2.2). They have a structure reminiscent of RLRs in which they contain

two CARD domains for signaling, a NOD domain, and a sequence of leucine rich repeats. They are maintained in a folded inactive state and upon binding to ATP and the PAMP at the NOD domain they undergo a conformational change to form a homodimer with other NLRs (Figure

2.5) (137).

Understanding the biological pathways and expression patterns of host PRRs has important implications for targeting. It allows the use of adjuvants and ligands in vaccine formulations to target specific receptors in order to achieve a tailored immune response. It is interesting to see repeated patterns within and throughout the families of PRRs. Examples include the abundance of leucine rich repeats on PRRs that recognize PAMPs and not nucleic acids as well as the prevalence of oligomerization for signal transduction. These observations can provide useful insights into the mechanisms that drive the immune response, which will ultimately allow us to design more efficacious vaccines.

2.2.3 Co-stimulation and crosstalk

Given the diverse repertoire of PAMPs on a pathogen it is likely that an infection will stimulate multiple PRRs, making crosstalk and co-signaling essential to combating infection. It may be important to replicate this in order to get an adequate immune response for vaccine delivery. There are many documented instances of PRR cooperation that center mostly around

TLRs. For example, TLRs have been shown to act synergistically with CLRs and to cooperate with NLRs. For example, Dectin-1 stimulation by β-glucans with simultaneous administration of

TLR2 or TLR4 agonists increased the pro-inflammatory cytokine secretion from primary monocytes and macrophages compared to the ligands being administered individually (138).

2.2.4 Targeting mechanisms

Numerous strategies have evolved for targeting of PRRs with an emphasis on generating potent immune responses. Targeting mechanisms are classified as passive or active. In passive targeting the drug does not have a target ligand and depends solely on its physicochemical properties as well as its interactions in the circulatory system or at tissue sites to be efficacious

(139). Active targeting employs a ligand that is conjugated to the drug allowing that drug to bind to a specific site on a cell or tissue (Figure 2.6). Active strategies are advantageous because they increase efficacy by achieving higher accumulation of the drug to the tissue, and are dose sparing, which can limit the drug exposure to non-targeted tissue (139). Active strategies that commonly target PRRs for drug delivery include functionalizing antibodies, functionalizing antigens, cell-mediated targeting, and functionalizing delivery vehicles.

2.2.4.1 Antibody-mediated targeting The strong binding and specificity of antibodies are ideal properties for delivering a payload to target cells. In the development of antibody-targeted vaccines, CLRs have been attractive targets for delivering antigens because of their endocytic capacity and their high expression levels on immature DCs. An antibody-conjugated vaccine that has undergone clinical trials is Lipovaxin-MM, produced by Lipotex, which is composed of a single domain antibody fragment that targets DC-SIGN for the delivery of a lipid-based vaccine that co-encapsulated melanoma antigens and IFN-γ (139). IFN-γ supplied the strong maturation signal that ensured that tumor-specific CD8+ T cells were activated (139). Separately, antibodies targeted to DC-

SIGN that had been conjugated with OVA generated persistent CD4+ and CD8+ T cell responses.

However, they still required a CD40 agonist to protect mice from OVA expressing Listeria monocytogenes.

In addition to CLRs, antibodies conjugated with antigen have also been targeted to other receptors commonly found on APCs. Fc receptors were among the first targets of antibody- conjugated vaccines. Instead of conjugating a payload to an entire antibody, the payload is usually conjugated to the Fc fragment, which is then recognized by the Fc receptor. Fc receptors are attractive targets because most of them contain ITAMs, which allow for a pro-inflammatory signal to be induced (140). Fc-conjugated drugs have demonstrated activation of CD4+ T cells and generally led to enhanced antigen-specific humoral and cellular immunity (141; 142).

Additionally, integrins such as CD11c have been targeted by antibody-conjugated drugs in order to ensure that antigens reach APCs (143). It has been reported that targeting through CD11c can induce T cell responses as effective as that by targeting some of the CLRs (144).

Antibody-mediated targeting has proven most successful in directly targeting the payloads to tumors. The vast majority of antibody-drug conjugates in clinical trials consist of traditional anticancer payloads, such as paclitaxel, docetaxel or doxorubicin encapsulated in liposomes or polymer-based nanomedicines, while the conjugated antibody fragments mediate targeting to the appropriate tissue (139). Receptors, such as the transferrin receptor or epidermal growth factor receptor are often overexpressed in malignant cells, making them excellent targets for antibodies (145; 146).

Given that antibodies are highly specific for their target and can easily be conjugated to deliver a payload, they are well suited for niche applications, such as targeting tumors or stimulating cell-mediated responses against HIV (147). Free cysteines on the protein provide flexibility for different drug conjugation techniques through a reaction with free sulfhydryls

(Table 2.3). However, their utility as a vaccine component against common respiratory infections is limited. For one, therapeutic antibodies currently on the market are expensive and one can

reasonably speculate that antibody-targeted vaccines would not be any more economical (148).

The antibodies themselves are fragile and require further modifications, such as PEGylation, to increase their half-life (139). Additionally, immune responses against the targeting antibody are possible, which induces rapid clearance and makes repeat immunizations unfeasible (149).

2.2.4.2 Cell-mediated targeting In cell-mediated targeting, immune cells are isolated from the blood of patients and exposed in vitro to an antigen or maturation stimuli, upon which they are re-injected into the patient. The process is labor intensive and costly, meaning that its use will most likely be restricted to highly specialized applications, such as cancer immunotherapy or treatment of HIV.

Two techniques are commonly used in cell-based targeting: DC immunotherapy or T cell adoptive transfers.

DC immunotherapy has demonstrated great proof of concept for use against a broad range of cancers. There are several studies in humans that demonstrate the safety and efficacy of this approach. In one study, patients with a form of liver cancer, known as hepatocellular carcinoma, had blood samples isolated from the peripheral blood, purified, and incubated with

GM-CSF and IL-4 to generate immature monocyte-derived DCs (MDDCs) (150). The MDDCs were incubated with lysate from a tumor cell line exposing the MDDCs to a number of antigens on the tumor. These patients were then vaccinated with the mature MDDCs, whereupon researchers witnessed a clinical response with increased IFN-γ secretion and decreasing serum levels of α-fetoprotein, a protein known to be expressed at elevated levels with this type of cancer (150). A major challenge of DC immunotherapy is ensuring that the DCs are sufficiently mature because evidence has shown that immature DCs could induce anergy rather than an antitumor immune response (150). A significant recent breakthrough is the use of TLR ligands to

induce MDDC maturation. Traditionally, a pro-inflammatory cytokine cocktail containing IL-6,

TNF-α, IL-1β, and Prostaglandin E2 had been used; however, studies have shown that these DCs do not do a very good job at producing IL-12p70 and thus do not induce Th1 responses as effectively (151). More antigen-specific T cell proliferation and higher surface expression of

CD83, CD86, and MHC II were observed upon incubating MDDCs with clinical grade Poly(I:C) or monophosphoryl lipid A (151; 152). Although a fairly new and developing field, DC immunotherapy has yet to live up to its promise. This is because generating de novo immune responses in advanced cancer patients is extremely difficult (153). Tumors progress because they promote an environment that induces immune suppression and immune dysfunction. As a result it is still unknown if DC immunotherapy primes T cells or if it simply activates pre-existing anergic T cells.

An alternative to DC immunotherapy is T cell adoptive immunotherapy. T cells that can penetrate into a tumor belong to a class of lymphocytes called tumor-infiltrating lymphocytes

(TILs). There is evidence showing that once the tumor-specific T cell meets its cognate antigen within the tumor, its migration is inhibited. The idea behind T cell adoptive immunotherapy is to excise tumor, extract tumor-specific T cells and encourage proliferation by adding pro- inflammatory growth factors such as IL-2 before re-injecting them into the patient (154).

2.2.4.3 Modified antigens as target agents A number of creative strategies have been used to modify antigens to actively target

APCs (Figure 2.6). Specific receptors on APCs are targeted allowing for the modified antigen to be internalized and presented more effectively compared to the unmodified antigen (98). Antigen modification techniques commonly include chemically conjugating either small molecules such as glycans or amino acids or full length antibodies to antigens (155; 156).

Perhaps the most common technique to modify antigens is to conjugate them with carbohydrates. Given that many antigens on pathogens are heavily glycosylated, researchers have attempted to use carbohydrate ligands for PRRs to create a “pathogen mimicking” antigen that triggers danger signals and initiates a strong immune response against the antigen (157).

Carbohydrates, such as mannose, fucose, galactose, β glucans, and Lewis X oligosaccharides have all been shown to be ligands for CLRs and are currently being exploited to modulate the immune response (158). In one study, researchers cultivated bone-marrow derived DCs

(BMDCs) and splenic DCs in vitro with OVA that had been modified with a DC-SIGN specific ligand, Lewis-X, using a bifunctional crosslinker known as 4-N-maleimidophenyl butyric acid

(Table 2.3) (158). Compared to unmodified OVA, the Lewis-X OVA enhanced antigen presentation, leading to greater T cell proliferation, particularly when OT-I T cells were added to the wells of the DCs. This suggests that these modified antigens have the ability to be cross presented, which is important for stimulating CD8+ T cell responses and inducing cell-mediated immunity (156; 158; 159).

In addition to conjugating carbohydrate ligands to antigens to target CLRs on APCs, TLR agonists and NLR ligands have also been studied (160-162). In one example, researchers chemically coupled CpG, a TLR9 agonist, to siRNA to target the TLR9+ myeloid cells and B cells (160). In turn the siRNA was designed to silence the immune suppressor gene Stat3 that would presumably hinder tolerogenic cells and stimulate a potent immune response in the presence of a tumor. An endosomal TLR, such as, TLR9 was targeted because it was believed that this would help the cell uptake the siRNA more efficiently. Incubation with mouse splenocytes proved that DCs, macrophages, and B cells efficiently internalized the conjugated siRNA more efficiently than the unconjugated siRNA. In vivo, local injection of the CpG-siRNA

oligonucleotide inhibited growth of colon carcinoma in mice and resulted in the up-regulation of co-stimulatory molecules such as MHC II, CD80, and CD40 on DCs and the production of Th1 cytokines (160). In this scenario, the CpG served two purposes in that it not only helped target the siRNA to immune cells, but also helped stimulate those cells in conjunction with the siRNA.

Targeting using TLR agonists has also caught the attention of industry and Novartis has used their TLR7 agonist to link an antigen from Streptococcus pneumoniae (163). They were able to show that mice administered 0.2 μg of the conjugated antigen had comparable survival rates as those given 2 μg of unconjugated antigen upon bacterial challenge (163). Mice given 0.2 μg of the unconjugated antigen with the soluble TLR7 agonist had significantly lower survival rates showing that chemical conjugation was important.

2.2.4.4 Modified delivery vehicles for active targeting Many of the same ligands and techniques to modify soluble antigens can also be used to modify delivery vehicles to prepare them for active targeting. Polymeric micro and nanoparticles make excellent delivery vehicles because of their ability to stabilize encapsulated protein, tailor release rates, become endocytosed by phagocytes, and enable high drug loading efficiency (164).

Recent research has shown that directly targeting CLRs on BMDCs leads to an enhanced immune response characterized by increased particle internalization, up-regulation of activation markers, such as MHC II, CD40, and CD86, as well as increased pro-inflammatory cytokine secretion (157; 165; 166). In these studies, polyanhydride nanoparticles were surface- functionalized with di-mannose in order to target the MMR and induce a pathogen-mimicking immune response. Functionalization was performed with a two-step reaction using 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide and N-hydroxysulfosuccinimide to link the carboxylic acids to a diamine linker (Table 2.3). It is known that pathogens such as Yersinia pestis and

Mycobacterium tuberculosis express conserved carbohydrate structures, including a variety of mannoses on their cell surface and so it is believed that including these sugars on the surface of particles will lead to a more efficacious nanovaccine (165). In addition to the sugars, there is evidence that polyanhydrides themselves act as TLR agonists, which lead to a more balanced cytokine response and a larger induction of CD8+ T cell response (167). The alternative to using sugars is to use antibodies specific for CLRs. One research group synthesized an antibody fusion protein targeted towards DEC-205 that was linked to chitosan microspheres for the intranasal delivery of plasmid DNA to mice (168). This resulted in larger quantities of mucosal IgA and as well as protein specific IgG when administered with an anti-CD40 DC maturation stimulus.

Instead of targeting CLRs, another approach is to surface functionalize biodegradable nanoparticles to target TLRs. This was performed by conjugating CpG nucleotides to the surface of PLGA nanoparticles in order to target TLR9 (169). Immunized mice had a strong Th1-biased immune response that included a high number of circulating effector T cells and ultimately resulted in a high degree of protection from West Nile virus infection (169). This makes sense because TLR9 is an endosomal receptor known for inducing more of a cell-mediated and Th1 polarized immune response when triggered (170). In the conjugation technique used, the CpG was chemically linked to biotin, while an avidin-linked fatty acid was incorporated on to the surface of the particles (Table 2.3) (171). Different variations of the biospecific biotin-avidin interactions are often used to link ligands to the surface of nanoparticles (168). Overall, conjugating TLR ligands to the surface of nanoparticles has been shown to adjuvant the immune response to minimize the need for antigen and to induce cellular immunity when targeting endosomal TLRs (172). Ligands for NLRs are more commonly encapsulated into particles as opposed to being surface functionalized with them. NOD1 and NOD2 ligands have been

encapsulated into PLA nanoparticles and co-delivered with encapsulated HIV Gag p24 antigen to increase the antibody response 100-fold compared to the antigen being administered with alum

(173). Encapsulating NOD ligands into nanoparticles will help APCs endocytose the ligands, but the ligands still have to escape the endosome and reach the cytosol to trigger the receptor.

2.2.5 Targeted vaccines for respiratory viruses

According to the World Health Organization, respiratory tract infections are the third leading cause of death in the world (174). There are many common respiratory viruses including influenza viruses, rhinoviruses, enteroviruses, respiratory syncytial viruses (RSV), and coronaviruses. An efficacious vaccine against these viral pathogens will result in a balanced immune response that activates a strong cell mediated response as well as induce an avid antibody humoral response. Innovative strategies are being used to design vaccines for respiratory viruses that specifically target PRRs. For example, there has been some interest in using flagellins, a TLR5 agonist, as an adjuvant to stimulate the immune response against influenza (175; 176). One group genetically fused the TLR5 ligand, flagellin, to the globular head of hemagglutinin derived from the influenza virus (177). Healthy human adults were immunized with the vaccine that ranged in dose from 0.1-8μg. Those receiving at least a 0.5μg dose of the vaccine generated strong hemagglutination-inhibiting antibody titers in excess of the

1:40 titer that is generally considered protective for influenza (177). The vaccine was deemed to be safe and generally was well tolerated. However, these subjects also tended to have strong antibody responses against the flagellin. Co-administering other TLR agonists with inactivated viruses or using protein subunits is becoming more common, such as the synthetically produced

TLR4 agonist, glucopyranosyl lipid A, as well TLR 3 agonists, such as Poly(I:C) (178; 179).

2.3 Influenza Virus and Influenza Vaccines

2.3.1 Introduction

During a given season, 5-15% of the world is infected with influenza resulting in 3-5 million instances of severe illness and nearly 500,000 deaths (180). Infection with the virus results in a highly infectious, acute respiratory disease with fever, headaches, malaise, and inflammation of the upper respiratory tract (181). Most influenza illnesses occur in children and in the elderly. This is because young children have yet to develop immunity against circulating strains, while aging has been associated with a decrease in immune function (182-184). In fact, nearly 90% of all influenza-related deaths are thought to occur in individuals older than 65 years of age (183). Additional complications can arise when the immune system is weakened by influenza allowing secondary infections, such as pneumonia or haemophilus influenzae to colonize the lower respiratory tract.

Accompanying the annual epidemics are pandemics that generally break out every 2-3 decades. A pandemic occurs when new viral epitopes are introduced into circulating strains, which means that the immune system has little to no memory to the newly circulating strain. The fact that all human pandemics in the last century have occurred due to avian or swine viral genes being introduced into the human population is exemplary of this (183). The most catastrophic influenza pandemic on record was the 1918 Spanish Flu that caused 50 million deaths worldwide, more fatalities than all of World War I (185). Two other pandemics occurred in the

20th century, including the Asian Flu of 1956 and the Hong Kong flu of 1967. The most recent pandemic was a new strain of the H1N1 virus, termed the swine flu, and it emerged out of

Mexico in 2009 (186). It spread rapidly to over 213 countries and resulted in at least 16,000 deaths. Researchers noticed a shift in the age distribution and mortality of the 2009 swine flu

pandemic as it afflicted younger populations more readily than the elderly. This is believed to be because the elderly may have had prior exposure to similar strains that had circulated decades earlier. This shift towards younger age groups is now known to be common in many pandemics and is considered to be a signature characteristic used to gauge the onset of a pandemic (187).

A rapid immune response is essential to limit the pathogenicity of these viruses, which is why vaccines that prime the immune system against these pathogens have helped limit the morbidity and spread of respiratory virus infections. To this end, in 1952, the World Health

Organization (WHO) established The Global Influenza Network (180). Since 1973, this organization has met before every flu season to determine emerging viral strains and provide recommendations so that manufacturers can produce vaccine strains that match circulating strains during winter outbreaks. This procedure aids in bringing over 250 million doses of inactivated vaccines to market annually (188). When the current vaccine is well matched to circulating strains, the vaccine has been shown to be effective in 70-90% of the population.

However, when the vaccine is poorly matched to the circulating strain, it has been shown to be no better than a placebo (180). Even with great leaps in medical technology throughout the last century, it is evident that influenza continues to cause widespread suffering and still results in a high degree of economic burden to society. As a result considerable improvement needs to be made in vaccine manufacturing and efficacy, and this starts with understanding the structure and pathogenicity of the influenza virus.

2.3.2 Structure and pathogenicity

The influenza virion consists of an enveloped lipid bilayer derived from its host and internally it has eight single strands of RNA that encode 11 proteins (Table 2.4) (185). Three of these proteins blanket the surface of the virion, which includes hemagglutinin (HA),

neuraminidase (NA) and the ion channel protein, M2 (Figure 2.7) (189). The remaining eight proteins are within the virion and function as a nucleocapsid protein, RNA polymerase complex, or an ion channel, or are non-structural proteins. Researchers have used the differences in the internal proteins to classify the virus into three species: A, B, and C. Influenza A infects humans and a number of mammals, including pigs, horses, wales, waterfowl, and chickens (189). Even though its host reservoir is in aquatic birds, this is the group that induced the aforementioned epidemics and pandemics in humans (183). Influenza B is restricted to humans, while Influenza

C can infect both humans and pigs. Both are less severe and less common than influenza A.

The influenza A virus is further classified by the differences in its surface proteins, HA and NA. Sixteen different HA (H1-H16) and nine different NA (N1-N9) protein subtypes exist

(183). Currently, only three HA serotypes, H1, H2, and H3, have become adapted to spread within the human population (190). Different serotypes of the influenza A virus have circulated throughout the human population in the last hundred years. Since the 1918 Spanish flu pandemic,

H1N1 was the primary virus in circulation throughout the world (191). In 1958, with the onset of the Asian flu pandemic, a new influenza virus with two new surface antigens appeared, known as

H2N2, and it completely displaced the H1N1 virus. This virus circulated for about 10 years until the Hong Kong flu pandemic in 1968, which was brought about by the H3N2 virus (191). This strain resulted from the avian H3 replacing the H2, while the NA remained the same. This new strain completely replaced the previous H2N2 virus in circulation. In 1978, the H1N1 virus reappeared and began circulating with the H3N2. In 2009, a new type of H1N1 virus emerged of swine origin and induced a pandemic. This new H1N1 has since replaced the original H1N1 and is now responsible for seasonal epidemics (191).

Once the virus enters the respiratory tract, the HA protein binds to the terminal sialic acid residues on epithelial cells. The cell then attempts to engulf the virus through endocytosis (192).

With a normal antigen, the cell’s endocytic vesicle would be gradually acidified, breaking the antigen down so that it can be displayed on the MHC molecules to activate the adaptive immune system. However, the virus hijacks this pathway as the acidic pH of the endosome activates the

M2 ion channel and allows acidification of the viral capsid (Figure 2.7) (183). This signals the

HA to fuse the viral membrane to the endosome followed by release of the viral ribonucleoproteins and the negatively stranded RNA into the cytoplasm of the host cell (185;

192). The virus must then rely on the cellular protein, importin, for each of the eight strands of

RNA to be imported into the nucleus. Once inside the nucleus, the negatively stranded RNA segments are transcribed by viral nucleoproteins to make messenger RNAs (mRNA). The mRNA is exported from the nucleus and uses the cell’s ribosomes to make new viral proteins.

These newly synthesized nucleoproteins are imported back into the nucleus where they meet up with viral genomic RNA and they are exported as a complex (185). Simultaneously, HA, NA and the M2 ion channel proteins are synthesized by the cell’s ribosomes and are inserted into the cell’s endoplasmic reticulum. They are then transported to the cell surface where the cytoplasmic tails of these three surface proteins interact with the matrix protein, M1, and the nucleocapsid proteins to finalize assembly of the virion. The newly formed virions bud from the cell’s plasma membrane ready to infect new cells. Newly budded virions tend to take on more of a filamentous shape, while viruses repeatedly passed in cell cultures for use as vaccines are typically spherical

(185). The NA protein plays a vital role in determining which cell the viral HA binds to as it can cleave the sialic acid residues preventing HA from binding to its target. The balance between HA binding to its ligand and NA cleaving the ligand interactions helps the virus reach its target cell.

Proper function of the HA protein is vital for an infectious virion. HA0 is a 70 kDa precursor of HA that is initially translated from cellular ribosomes and is on the surface of newly formed virions (190). It assembles as a trimer on the surface of the virus, which increases the avidity to host cells through multivalent binding (Figure 2.7). In the bronchiolar epithelium, the

HA0 trimer encounters cellular serine proteases, such as trypsin, which cleave it into two subunits, HA1 and HA2 (190; 193). HA2 is anchored to the viral envelope while HA1 is held to

HA2 through disulfide interactions (194). The HA1 subunit contains the sialic acid binding domain that allows the virus to be taken up into a cell’s endosome. It also contains the globular head where most of the epitopes for stimulating an immune response are located (183). As the endosome acidifies the entire protein undergoes an irreversible and thermodynamically favorable conformational change exposing a fusion peptide domain on the HA2 subunit (193). This peptide domain is hydrophobic allowing for insertion into the cell membrane, thus bringing the cell endosome and the viral membrane close together for fusion. For most HAs, the cleavage of specific proteases limits infection to only the lungs; however, the avian flu, H5, is more worrisome because its HA0 precursor is able to be cleaved by a broader range of proteases, such as furin, meaning the virus can spread to other organs, ultimately inducing higher virulence

(190).

To optimize its pathogenicity, the influenza virus has mechanisms in place to suppress the host immune system. This is mostly done through nonstructural protein 1 (NS1), which has been shown to bind and inhibit RLR signaling (Table 2.4) (195; 196). Type I interferon (IFN-I) expression is reduced in the host and there is little release of pro-inflammatory cytokines (195;

197). Viruses lacking the NS1 protein have been shown to induce significant IFN-I and inhibit viral replication (195). Furthermore, NS1 outcompetes the host antiviral protein, double-stranded

RNA-dependent protein kinase (PKR) for double stranded RNAs to suppress its activity (195).

Downstream PKR pathways lead to inhibition of cellular translation, meaning viral proteins cannot be synthesized from its mRNAs. It also activates the transcription factor, Interferon regulatory factor 3 (IRF3), which leads to the expression of IFNs. Mounting evidence suggests that the NP protein also has its own unique role in inhibiting PKR activation (198). Lastly, NS1 has also been shown to activate phosphatidylinositol 3-kinase (PI3K) in the host, which is a pathway that suppresses apoptosis allowing the virus uninterrupted replication and production of new virions (197).

In addition to suppressing the host immune system, the virus is able to directly evade the immune system through two robust strategies, antigenic drift and antigenic shift. The former occurs because the influenza virus doesn’t have any “proofreading” proteins to correct errors made by the RNA polymerase when it synthesizes RNA (180). This is evolutionarily advantageous, because it ensures that mutations will be common, resulting in human antibodies no longer being able to opsonize the virus. As a result, vaccines need to be updated every year so that they can match the strains of circulating viruses. Most antigenic drift sites are located in the globular head of the HA1 subunit as the HA2 region has been shown to be strongly conserved

(183). This makes sense as the globular head of the HA is exposed to antibody opsonization, while the HA2 subunit remains mostly hidden and has important functions in fusion, which could mutate and result in lower efficacy (Figure 2.7). Even more worrisome is that researchers have shown that as few as two amino acid mutations on the H3 or H1 protein can alter its specificity for human or avian receptors (190). It is believed that the 1918 Spanish flu pandemic occurred because of mutations from the avian H1N1 virus allowing it to infect humans (185).

Antigenic shift occurs when two different types of flu viruses infect a cell and exchange

RNA genes in a process called gene reassortment. This is only able to happen because the influenza genome consists of eight individual segments and so if the host is simultaneously infected with two viruses, the progeny virions could be composed of RNA segments from both of the original virions. Reassortment has significant potential to cause pandemics as a new HA subtype can easily be introduced into a population where there has been no prior exposure to that

HA to gain immunity (185). The 1957 Asian flu pandemic, the 1968 Hong Kong flu, and the

2009 swine flu pandemic were all caused by reassortment viruses. Pigs are often highlighted due to their susceptibility to serve as mixing vessel for reassortment (183). Specifically, the upper respiratory tract of humans is composed of sialic acid residues bonded with an α2-6 linkage to galactose, while the respiratory tract of avian species is mostly composed of sialic acid residues bonded with an α2-3 linkage to galactose (199). HA on influenza viruses specific to avian species can only bind to the α2-3 terminated sialic acids, while HA specific to humans can only bind to α2-6 terminated sialic acids (199). Pigs on the other hand have α2-3 and α2-6 linked sialic acids, which makes them vulnerable to both avian and human viruses. Moreover, this provides an enhanced opportunity for avian associated genes to be incorporated into influenza viruses specific to humans through reassortment (199). In regards to the avian H5N1 virus, this is a serious concern. The H5N1 virus is highly pathogenic and infection results in >70% mortality rate and yet the virus is poorly transmitted in humans (200). In a worst case scenario, the virus would acquire the ability to be efficiently transmitted, perhaps through reassortment, which could induce a potential pandemic due to the lack of adaptive immunological memory in the human population.

Current circulating strains of the influenza virus are spread through small particle aerosols expelled through the air of infected individuals who are coughing (201). This is how the virus maintains itself in the human population. Influenza viruses induce symptoms that last 7-10 days and viral replication generally peaks at 48 hours after infection, but continues to be shed for

6-7 days (181). The virus initially infects through the upper respiratory tract, but in later stages replicates in both the upper and lower respiratory tract. Viral infection in the upper respiratory tract causes a loss of the ciliated epithelium, which hinders the host’s ability to trap pathogens in mucus and stops mucociliary flow. Replication triggers the release of pro-inflammatory cytokines and interferons, which can lead to a systemic immune response that results in the classical flu symptoms (185). Serious complications and secondary infections are less common, but can occur in high risk groups, such as those who are immunocompromised, those with chronic pulmonary diseases, or the elderly. Influenza can be diagnosed by analyzing viral antigens or genetic material in clinical samples, or by monitoring the rise and fall in specific antibody titers. A very sensitive test is reverse transcriptase-polymerase chain reaction (RT-PCR) to look for evidence of the viral genome in respiratory specimens (202). Serologic detection for viral antigens is less common.

The best way for the host to limit the influenza virus is to target the HA protein with neutralizing antibodies because this can prevent the virus from binding and fusing to host cells

(6). Antibodies targeting the NA protein help reduce viral replication by inhibiting the release of new viral particles; however they cannot neutralize infectivity. The influenza virus is recognized by three PRRs that help stimulate the immune system. Its single-stranded RNA is recognized by

TLR7 and RIG-I, while the matrix protein M2 activates a member of the NLR family called

NLRP3 (203). Activation of TLR7 has been found to help antibody isotype switching directly on

B cells. RIG-I has a redundant role in that it also recognizes viral RNA, but it is expressed on a more diverse repertoire of cells including epithelial cells and mast cells, both of which up- regulate pro-inflammatory cytokines and Type I interferon (203). Viral clearance from the host also depends on an adaptive cell-mediated immune response. Cytotoxic T lymphocytes are known to have epitopes to some of internal proteins of the virus including NP, PB2, and PA

(183; 204). A robust cell-mediated immune response cannot prevent infection against influenza, but it can prevent illness and even death. Because the internal proteins of the influenza virus are more conserved, a cell-mediated immune response is more apt to handle evolving viral strains.

2.3.3 H1N1 Virus

The H1N1 virus is currently one of two circulating influenza A viruses in the human population with the other being the H3N2 virus. The H1N1 virus is most notorious for its role in both the 1918 Spanish flu pandemic and the 2009 swine flu pandemic. Additional periodic outbreaks of H1N1 in the past have also generated widespread fears of an H1N1 pandemic, which includes the outbreak at an American military base on Fort Dix in 1976 and the Russian

Flu epidemic of 1977. More specifically, during the winter of 1975-1976, 200 military personal on Fort Dix fell ill with the H1N1 virus. Previously the H1N1 virus of avian origin was introduced into the human population through genetic drift in 1918 and circulated until 1958 when it was replaced by the H2N2 virus during the Asian flu pandemic. The H1N1 virus was not in human circulation during the Fort Dix outbreak and so it was concluded that the outbreak was caused by zoonotic transmission of swine H1N1 virus into a population that lived in close quarters (205). A proactive response was adopted by the military that ensured that the virus never made it off the base and into the general population. Zoonotic transmission of avian or swine viruses are somewhat common for those who work closely with the animals, but they generally

do not result in widespread human-to-human transmission without genetic changes. The H1N1 virus reappeared once more in 1978 originating from Russia. It was similar to the virus that circulated until the 1950’s meaning that those under 25 years of age had no immunological memory of the virus and had the highest rate of infection (206).

Phylogenetic analysis revealed that the 2009 swine influenza virus received its HA, NA, and NP proteins from a triple reassortment virus that had been circulating in swine (207). This reassortment virus itself had RNA segments that originated from avian, swine, and human viruses. In addition, the 2009 swine influenza virus received its RNA that encoded the NA and matrix proteins from a Eurasian swine virus (207). This virus had a unique gene composition allowing it to quickly circulate in humans around the globe and making it the first pandemic of the 21st century (208). Infection from the virus was relatively mild compared to other pandemics.

This is because the virus didn’t induce a strong cytokine storm like the 1918 H1N1 virus or the

H5N1 virus that would cause strong damage in the lungs of the host. Researchers found that it was the NA protein that induced the hyper cytokine production in H5N1 and the fear was that if the H1N1 virus acquired this gene it would result in a virus that could cause many fatalities

(209). Even though the virus was milder than what was originally feared, the world faced significant shortages of the vaccine during 2009 due to manufacturing problems. This left millions of people unvaccinated. Adjuvants were not used in the vaccine, but if they had been, experts estimate that it would have allowed four times as many people to be immunized leaving less people exposed (210).

2.3.4 Traditional vaccines and new vaccination approaches

The severity of influenza infections is most effectively prevented by vaccination or through the use of antiviral drugs. Vaccination is the best prophylactic technique to prevent

infection, while antivirals are good therapeutics. Additionally, when a vaccine is given to an entire population, it can induce herd immunity, which will prohibit the virus from spreading throughout the population (183).

Current seasonal vaccines depend on generating an antibody response to the HA proteins.

Vaccines are formulated from the H1N1, H3N2, and an Influenza B strain, and they are either inactivated viruses or live, attenuated viruses (183). In both forms, the virus is generally mass produced by propagation in chicken eggs. The selected influenza A strain is co-infected into chicken eggs with a laboratory adapted strain (PR8) designed for rapid growth in eggs. These newly “reassorted” viruses are selected for their ability to grow rapidly and still have the epidemic surface proteins (188). It is this strain that is distributed to manufacturers for propagation in eggs; however, for influenza B, there are no high growth laboratory adapted strains and so the mass produced vaccine is from the wild-type virus. In addition, since 2013, the

WHO has included a second influenza B strain in their annual list of strain recommendations and so some manufacturers have decided to make a quadrivalent vaccine (188). The virus and the fluid of the egg, the allantoic fluid, are harvested. Centrifugation at extremely high speeds removes the virus from soluble contaminants. After processing, one egg yields enough virus for approximately three doses of vaccine for one strain, while the total manufacturing process for a season’s worth of vaccines can take up to six months (183; 211).

Only since 2007 have vaccines become available from viruses cultured in cells (188;

212). By avoiding eggs, manufacturers don’t have to worry about egg allergies within the general public, and can better control scale-up and sterility. The most successful cell lines for growing up the influenza virus are Vero cells and Madin-Darby Canine Kidney (MDCK) cells (212).

However, this process is still very technically challenging and so only a small fraction of

vaccines are produced through this technique. For example, the surface proteins HA and NA are heavily glycosylated and one challenge with viruses produced in cell culture is replicating these glycosylation patterns. In a recent study, researchers found that HA derived from viruses grown in MDCK cells had complex glycans that were terminated in α or β galactoses (213). In contrast,

HA derived from viruses in Vero cells had low molecular weight glycans that terminated only in

β-galactose or high mannoses. The latter variant stimulated splenocytes to produce more IL-2 and IFN-γ, while enhancing T cell proliferation (213). It is hypothesized that the HA produced from Vero cells is more immunogenic, because the high mannose glycans were able to interact with CLRs.

After culturing the virus in cells or eggs, the grown influenza virus is then chemically treated and combined with other strains to make a trivalent inactivated vaccine (TIV). The chemical treatment method used will determine the specific type of vaccine. Whole inactivated vaccines are made by treating the virus with formaldehyde or β-propiolactone. Split vaccines and subunit vaccines are the result of viruses being treated with sodium dodecyl sulfate or other detergents. The detergents breakdown the viral membrane releasing all proteins into solution

(188). In the subunit vaccine, the HA protein is then purified, whereas all of the contents remain in the split vaccine. The subunit vaccine is known to induce the least amount of adverse reactions, but is also the least immunogenic (183). To compensate for the lack of immunogenicity it is often times given in two administrations to young children. Typically TIV’s consist of 15 μg of each HA antigen for a total of 45 μg of HA protein delivered. For the elderly, the dosage has been increased to 60 μg of each HA (214). It is well known that aging is associated with a decline in the immune system, which means a decline in vaccine efficacy

(182). Researchers have found that one reason for this is because TLR expression is decreased

and dysregulated on APCs of older individuals. APCs from young adults were able to produce statistically higher amounts of TNF-α, IL-6, and IL-12 after TLR engagement compared to older adults (182).

One problem with the TIV vaccines is that they do not elicit a strong mucosal or cytotoxic T lymphocyte immune response (215). In contrast live-attenuated influenza vaccines

(LAIVs), administered intranasally, have been shown to induce strong IgA antibodies in the respiratory tract (216). Given that the influenza virus is a respiratory infection this is a very important consideration as an efficacious mucosal response could stop the virus from infecting the host. The viruses in the LAIVs are reassortment viruses, produced by co-infecting cells or chicken eggs with a virus strain that has been passed through low temperatures and only replicates efficiently at 25°C along with the epidemic viruses (189). The temperature in the upper respiratory tract is slightly higher and so the attenuated virus cannot replicate efficiently and has reduced virulence. Yet it is still able to efficiently mimic a natural infection that helps prime the immune system. LAIVs are produced by MedImmune and have been produced as either a trivalent or quadrivalent vaccine. Numerous studies have tried to directly compare LAIVs to

TIVs. In clinical trials with young children, LAIVs have been shown to be more efficacious and they resulted in reduced illness than TIVs (217-219). Fewer comparative studies exist for adults; however, there was one clinical study involving nearly 2000 healthy young adults that showed the TIV vaccine resulted in the best protection (220). This could potentially be because the TIVs are known to induce a strong systemic response that leads to a better serum antibody response compared to LAIVs. The general consensus in adults is that TIVs induce as good or a better immune response than LAIVs (221). Given that LAIVs are live viruses, there is some concern that these viruses could revert to their pathogenic form in immunocompromised individuals. One

clinical study showed that this generally was not a problem because no adverse reactions occurred and there was no detection of increased viral shedding (222).

A new generation of novel vaccines for influenza is in development, which is working to overcome the limitations of TIVs and LAIVs (223). These newer vaccines do not use eggs to grow the virus, but often use expression systems to recombinantly produce HA subunits. The most successful of these is the baculovirus expression system (BEVS). This protocol was first used to make the papilloma virus vaccine and given its record of safety and ability for mass production is now being extended to influenza (224; 225). Baculoviruses are DNA viruses that have the HA coded into them. The baculovirus is able to infect insect cells, which makes large amounts of protein that has been coded into the viral genome. In the case of HA, the system makes the HA0 protein that doesn’t get cleaved into HA1 or HA2. Experiments with recombinant H5 proteins that were made from both insect cells and mammalian Chinese Hamster

Ovary (CHO) cells showed that the H5 produced from both cell lines had undergone post- translation modifications and were glycosylated differently (226). Insect-produced H5 had N- glycans consisting of mannose that were significantly shorter and less complicated than the mammalian cells lines. Upon in vivo injection into mice, the H5 produced from insect cell lines had the highest antibody titers. However, the H5 from mammalian cells lines had better neutralizing and HA inhibition titers allowing mice immunized with this H5 to have a better chance at protective immunity when challenged with a live virus (226). This is another example illustrating the importance of how recombinantly produced proteins can be affected by their host cell line. Yet other studies have shown that HA produced using the BEVS can induce the same immunogenicity as egg based TIVs (227). In terms of mass production, this system also has the advantage that no live virus is used, lowering the danger level for those involved in the

production process, unlike the egg-based or cell culture methods. One company, Protein

Sciences Corporation, has an FDA-approved influenza vaccine, named FluBlok, that was generated using insect cell fermentations (228). The entire process of cloning, expression, and manufacturing took two months, which is significantly shorter than the six months that is generally needed for the egg-based method (227; 229).

There has also been interest in using plants as a source to recombinantly express HA antigens. Plant expression systems would have the same beneficial safety features for mass production as the BEVS and the potential for scalability (230). In the plant system, a viral vector named Agrobacterium tumefaciens is used to introduce recombinant proteins into tobacco plants.

Plant produced HA is antigenic as confirmed by ELISAs and has been shown to induce an immune response in mice with as little as 5 μg (231). Studies with mice and ferrets showed that they induce both cellular and humoral immune responses (231; 232). Usually the antigen is administered in conjunction with adjuvants, such as, Quil-A or Alum (230).

Delivery systems are being designed to enhance the versatility of subunit proteins produced through these expression systems and adjuvant the immune response. Using polyanhydride delivery vehicles, researchers have demonstrated that H5 antigen encapsulation into nanoparticles results in sustained protein release and enhanced protein stability during in vitro degradation studies (55). Whereas soluble antigen results in weaker immune responses requiring larger doses, the polyanhydride nanoparticles slowly release stable and functional proteins while also serving as an adjuvant to provide a dose sparing effect for expensive proteins

(60). Other types of polymers have been used to synthesize nanoparticles and encapsulate or adsorb influenza antigens to their surface. This includes the use of calcium phosphate, PCL, or

PLGA nanoparticles (72; 227; 233; 234). Antigen on calcium phosphate nanoparticles is

functionalized to the surface because the calcium phosphate is a hard inorganic material that only breaks down in the lysosomes of cells (233). HA loaded PCL particles yielded strong hemagglutinin inhibition titers and induced a more balanced Th1 and Th2 immune response with a single intranasal or intramuscular immunization in mice (72). Using nanoparticle delivery systems opens up new routes for vaccine delivery including intranasal and oral delivery because the antigen is stabilized and is no longer exposed to a degradative environment. In one example, researchers found that subunit antigen had low mucosal permeability and was often times unstable in the intestinal environment (235). Encapsulating the antigen into nanoparticles stabilized the antigen and the particles were easily phagocytosed by M cells (235).

The poor immunogenecitiy of subunit vaccines can be overcome by incorporating adjuvants into vaccine formulations, which can be used with or without delivery vehicles.

Adjuvants are compounds that increases the potency of an immune response to an antigen by promoting antigen uptake or targeting specific immune pathways (211). Adjuvants that have been FDA-approved for vaccines include the oil-in-water emulsions MF59 and AS03, aluminum salts known as alum, and monophosphoryl lipid A (MPLA) (183; 236). MPLA is approved for use in the human papilloma virus vaccine and the Hepatitis B virus vaccine, but is not often found in influenza vaccines. Alum is sometimes used in influenza TIV vaccines and it functions by activating the NLRP3 complex as part of the NLR family of PRRs. Antigen adsorbed to alum is presented as a particulate allowing for better internalization and ultimately induction of a strong antibody response (236). Alum has a strong safety record since the 1920’s, but the disadvantage is that it does not elicit a strong cell-mediated immune response (237). A more recent alternative is MF59, which is produced by Novartis. MF59 has been used in many studies with influenza and although its mechanism is unknown, the adjuvant has been shown to induce

long lasting antibody and T cell responses (238; 239). It works similarly to alum by enhancing antigen uptake by APCs and specifically it has been shown to be more effective than alum at recruiting neutrophils to the injection site and inducing better antigen transport to the draining lymph nodes (240). Even though it induces a strong immune response at the injection site, unlike alum, it does not establish an antigen depot (241). AS03 is also an oil-in-water emulsion made by

GlaxoSmithKline and it has been approved for use against H5N1 and was extensively tested in the 2009 H1N1 pandemic influenza vaccine (242; 243).

Another delivery system being explored as vaccine adjuvants and/or delivery vehicles are virus-like particles (VLPs), which are essentially an assembly of influenza structural proteins.

Morphologically they look like a live virus, but they are packaged without RNA and so they are unable to replicate, making them safer than live attenuated vaccines (183). VLPs have been produced in baculovirus and plant expression systems (224; 244). The plant-produced VLPs in combination with alum have been able to induce an immune response in mice with as little as 0.5

μg (230). The baculovirus expression system was used to produce H9N2 VLPs, which expressed

HA, NA, and M1 (245). These recombinant structural proteins were able to self-assemble into

80-120 nm diameter spheres when expressed and induced protective immune responses in

BALB/c mice. It has been shown that VLPs induce an immune response with a different profile compared to soluble recombinant HA as not only are antibodies against NA produced, but there is also more class switching to a Th1 based immune response (246).

2.4 Conclusions Vaccines have a long and proven record of positively contributing to public health. The use of nanotechnologies as a delivery platform as well as incorporating ligands that actively target immune cells will further improve the efficacy of future vaccines. These next generation

systems will have the potential to provide controlled antigen release, dose sparing capabilities, single dose vaccination, and modulation of the immune response. Many of these next generation nanovaccines are already in development to combat the influenza virus. The following chapter introduces the specific research objectives for this thesis. Later chapters describe research that utilizes surface functionalized polyanhydride nanovaccines in order to actively target immune cells and future work will investigate applying these systems to design an influenza vaccine.

2.5 References 1. De Gregorio E, Rappuoli R. 2014. From empiricism to rational design: a personal perspective of the evolution of vaccine development. Nature reviews. Immunology 14:505-14

2. Kumari A, Yadav SK, Yadav SC. 2010. Biodegradable polymeric nanoparticles based drug delivery systems. Colloid Surface B 75:1-18

3. Seong SY, Matzinger P. 2004. Hydrophobicity: an ancient damage-associated molecular pattern that initiates innate immune responses. Nature reviews. Immunology 4:469-78

4. van Vlerken LE, Vyas TK, Amiji MM. 2007. Poly(ethylene glycol)-modified nanocarriers for tumor-targeted and intracellular delivery. Pharm Res 24:1405-14

5. He C, Hu Y, Yin L, Tang C, Yin C. 2010. Effects of particle size and surface charge on cellular uptake and biodistribution of polymeric nanoparticles. Biomaterials 31:3657-66

6. Bachmann MF, Jennings GT. 2010. Vaccine delivery: a matter of size, geometry, kinetics and molecular patterns. Nature reviews. Immunology 10:787-96

7. Liu ZH, Jiao YP, Wang YF, Zhou CR, Zhang ZY. 2008. Polysaccharides-based nanoparticles as drug delivery systems. Adv Drug Deliver Rev 60:1650-62

8. Nobs L, Buchegger F, Gurny R, Allemann E. 2004. Current methods for attaching targeting ligands to liposomes and nanoparticles. Journal of pharmaceutical sciences 93:1980-92

9. Patri AK, Kukowska-Latallo JF, Baker JR, Jr. 2005. Targeted drug delivery with dendrimers: comparison of the release kinetics of covalently conjugated drug and non-covalent drug inclusion complex. Adv Drug Deliv Rev 57:2203-14

10. Kocbek P, Obermajer N, Cegnar M, Kos J, Kristl J. 2007. Targeting cancer cells using PLGA nanoparticles surface modified with monoclonal antibody. Journal of Controlled Release 120:18-26

11. Jiskoot W, Randolph TW, Volkin DB, Middaugh CR, Schoneich C, et al. 2012. Protein instability and immunogenicity: roadblocks to clinical application of injectable protein delivery systems for sustained release. Journal of pharmaceutical sciences 101:946-54

12. Petersen LK, Huntimer L, Walz K, Ramer-Tait A, Wannemuehler MJ, Narasimhan B. 2013. Combinatorial evaluation of in vivo distribution of polyanhydride particle-based platforms for vaccine delivery. International journal of nanomedicine 8:2213-25

13. Danhier F, Ansorena E, Silva JM, Coco R, Le Breton A, Preat V. 2012. PLGA-based nanoparticles: an overview of biomedical applications. Journal of controlled release : official journal of the Controlled Release Society 161:505-22

14. Lopez-Cebral R, Martin-Pastor M, Seijo B, Sanchez A. 2014. Progress in the characterization of bio-functionalized nanoparticles using NMR methods and their applications as MRI contrast agents. Progress in nuclear magnetic resonance spectroscopy 79:1-13

15. Moghimi SM, Hunter AC, Murray JC. 2005. Nanomedicine: current status and future prospects. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 19:311-30

16. Kawasaki ES, Player A. 2005. Nanotechnology, nanomedicine, and the development of new, effective therapies for cancer. Nanomedicine : nanotechnology, biology, and medicine 1:101-9

17. Chasin M, Langer R, eds. 1990. Biodegradable Polymers as Drug Delivery Systems. New York, NY: Marcel Dekker, Inc. 347 pp.

18. Jain RA. 2000. The manufacturing techniques of various drug loaded biodegradable poly(lactide-co-glycolide) (PLGA) devices. Biomaterials 21:2475-90

19. Makadia HK, Siegel SJ. 2011. Poly Lactic-co-Glycolic Acid (PLGA) as Biodegradable Controlled Drug Delivery Carrier. Polymers-Basel 3:1377-97

20. Wu XS, Wang N. 2001. Synthesis, characterization, biodegradation, and drug delivery application of biodegradable lactic/glycolic acid polymers. Part II: Biodegradation. J Biomat Sci-Polym E 12:21-34

21. Zolnik BS, Burgess DJ. 2007. Effect of acidic pH on PLGA microsphere degradation and release. Journal of controlled release : official journal of the Controlled Release Society 122:338-44

22. Soppimath KS, Aminabhavi TM, Kulkarni AR, Rudzinski WE. 2001. Biodegradable polymeric nanoparticles as drug delivery devices. Journal of Controlled Release 70:1-20

23. Mundargi RC, Babu VR, Rangaswamy V, Patel P, Aminabhavi TM. 2008. Nano/micro technologies for delivering macromolecular therapeutics using poly(D,L-lactide-co-glycolide) and its derivatives. Journal of controlled release : official journal of the Controlled Release Society 125:193-209

24. Qaddoumi MG, Gukasyan HJ, Davda J, Labhasetwar V, Kim KJ, Lee VH. 2003. Clathrin and caveolin-1 expression in primary pigmented rabbit conjunctival epithelial cells: role in PLGA nanoparticle endocytosis. Molecular vision 9:559-68

25. Harush-Frenkel O, Debotton N, Benita S, Altschuler Y. 2007. Targeting of nanoparticles to the clathrin-mediated endocytic pathway. Biochemical and biophysical research communications 353:26-32

26. Cartiera MS, Johnson KM, Rajendran V, Caplan MJ, Saltzman WM. 2009. The uptake and intracellular fate of PLGA nanoparticles in epithelial cells. Biomaterials 30:2790-8

27. Shen H, Ackerman AL, Cody V, Giodini A, Hinson ER, et al. 2006. Enhanced and prolonged cross-presentation following endosomal escape of exogenous antigens encapsulated in biodegradable nanoparticles. Immunology 117:78-88

28. Hamdy S, Molavi O, Ma ZS, Haddadi A, Alshamsan A, et al. 2008. Co-delivery of cancer-associated antigen and Toll-like receptor 4 ligand in PLGA nanoparticles induces potent CD8(+) T cell-mediated anti-tumor immunity. Vaccine 26:5046-57

29. Murata N, Takashima Y, Toyoshima K, Yamamoto M, Okada H. 2008. Anti-tumor effects of anti-VEGF siRNA encapsulated with PLGA microspheres in mice. Journal of controlled release : official journal of the Controlled Release Society 126:246-54

30. Patil Y, Panyam J. 2009. Polymeric nanoparticles for siRNA delivery and gene silencing. International journal of pharmaceutics 367:195-203

31. Tian JY, Yu J. 2011. poly (lactic-co-glycolic acid) nanoparticles as candidate DNA vaccine carrier for oral immunization of Japanese flounder (Paralichthys olivaceus) against lymphocystis disease virus (vol 30, pg 109, 2011). Fish Shellfish Immun 31:364-

32. van de Weert M, Hennink WE, Jiskoot W. 2000. Protein instability in poly(lactic-co- glycolic acid) microparticles. Pharmaceut Res 17:1159-67

33. Fu K, Pack DW, Klibanov AM, Langer R. 2000. Visual evidence of acidic environment within degrading poly(lactic-co-glycolic acid) (PLGA) microspheres. Pharmaceut Res 17:100-6

34. Chawla JS, Amiji MM. 2002. Biodegradable poly(epsilon -caprolactone) nanoparticles for tumor-targeted delivery of tamoxifen. International journal of pharmaceutics 249:127-38

35. Jameela SR, Suma N, Jayakrishnan A. 1997. Protein release from poly(epsilon- caprolactone) microspheres prepared by melt encapsulation and solvent evaporation techniques: a comparative study. Journal of biomaterials science. Polymer edition 8:457-66

36. Sinha VR, Bansal K, Kaushik R, Kumria R, Trehan A. 2004. Poly-epsilon-caprolactone microspheres and nanospheres: an overview. International journal of pharmaceutics 278:1-23

37. Kumar N, Langer RS, Domb AJ. 2002. Polyanhydrides: an overview. Adv Drug Deliv Rev 54:889-910

38. Narasimhan B, Kipper MJ. 2004. Surface-Erodible Biomaterials for Drug Delivery. In Advances in Chemical Engineering, 29:169-218: Elsevier, Inc. Number of 169-218 pp.

39. Shen E, Pizsczek R, Dziadul B, Narasimhan B. 2001. Microphase separation in bioerodible copolymers for drug delivery. Biomaterials 22:201-10

40. Torres MP, Determan AS, Mallapragada S, Narasimhan B. 2006. Polyanhydrides. In Encyclopedia of Chemical Processing:2247-57. Taylor and Francis. Number of 2247-57 pp.

41. Katti DS, Lakshmi S, Langer R, Laurencin CT. 2002. Toxicity, biodegradation and elimination of polyanhydrides. Adv Drug Deliv Rev 54:933-61

42. Petersen LK, Sackett CK, Narasimhan B. 2010. High-throughput analysis of protein stability in polyanhydride nanoparticles. Acta biomaterialia 6:3873-81

43. Kipper MJ, Shen E, Determan A, Narasimhan B. 2002. Design of an injectable system based on bioerodible polyanhydride microspheres for sustained drug delivery. Biomaterials 23:4405-12

44. Gopferich A, Tessmar J. 2002. Polyanhydride degradation and erosion. Adv Drug Deliv Rev 54:911-31

45. Determan AS, Wilson JH, Kipper MJ, Wannemuehler MJ, Narasimhan B. 2006. Protein stability in the presence of polymer degradation products: consequences for controlled release formulations. Biomaterials 27:3312-20

46. Dang WB, Daviau T, Ying P, Zhao Y, Nowotnik D, et al. 1996. Effects of GLIADEL(R) wafer initial molecular weight on the erosion of wafer and release of BCNU. Journal of Controlled Release 42:83-92

47. Perry J, Chambers A, Spithoff K, Laperriere N. 2007. Gliadel wafers in the treatment of malignant glioma: a systematic review. Current oncology 14:189-94

48. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

49. Determan AS, Graham JR, Pfeiffer KA, Narasimhan B. 2006. The role of microsphere fabrication methods on the stability and release kinetics of ovalbumin encapsulated in polyanhydride microspheres. Journal of microencapsulation 23:832-43

50. Carrillo-Conde B, Schiltz E, Yu J, Chris Minion F, Phillips GJ, et al. 2010. Encapsulation into amphiphilic polyanhydride microparticles stabilizes Yersinia pestis antigens. Acta biomaterialia 6:3110-9

51. Torres MP, Determan AS, Anderson GL, Mallapragada SK, Narasimhan B. 2007. Amphiphilic polyanhydrides for protein stabilization and release. Biomaterials 28:108-16

52. Determan AS, Trewyn BG, Lin VS, Nilsen-Hamilton M, Narasimhan B. 2004. Encapsulation, stabilization, and release of BSA-FITC from polyanhydride microspheres. Journal of controlled release : official journal of the Controlled Release Society 100:97-109

53. Petersen LK, Phanse Y, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2012. Amphiphilic polyanhydride nanoparticles stabilize Bacillus anthracis protective antigen. Molecular pharmaceutics 9:874-82

54. Haughney SL, Petersen LK, Schoofs AD, Ramer-Tait AE, King JD, et al. 2013. Retention of structure, antigenicity, and biological function of pneumococcal surface protein A (PspA) released from polyanhydride nanoparticles. Acta biomaterialia 9:8262-71

55. Ross KA, Loyd H, Wu W, Huntimer L, Wannemuehler MJ, et al. 2014. Structural and antigenic stability of H5N1 hemagglutinin trimer upon release from polyanhydride nanoparticles. Journal of biomedical materials research. Part A

56. Vela Ramirez JE, Roychoudhury R, Habte HH, Cho MW, Pohl NL, Narasimhan B. 2014. Carbohydrate-functionalized nanovaccines preserve HIV-1 antigen stability and activate antigen presenting cells. Journal of biomaterials science. Polymer edition 25:1387-406

57. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

58. Kipper MJ, Wilson JH, Wannemuehler MJ, Narasimhan B. 2006. Single dose vaccine based on biodegradable polyanhydride microspheres can modulate immune response mechanism. Journal of biomedical materials research. Part A 76:798-810

59. Ulery BD, Kumar D, Ramer-Tait AE, Metzger DW, Wannemuehler MJ, Narasimhan B. 2011. Design of a protective single-dose intranasal nanoparticle-based vaccine platform for respiratory infectious diseases. PloS one 6:e17642

60. Huntimer L, Wilson Welder JH, Ross K, Carrillo-Conde B, Pruisner L, et al. 2013. Single immunization with a suboptimal antigen dose encapsulated into polyanhydride microparticles promotes high titer and avid antibody responses. Journal of biomedical materials research. Part B, Applied biomaterials 101:91-8

61. Ross KA, Haughney SL, Petersen LK, Boggiatto P, Wannemuehler MJ, Narasimhan B. 2014. Lung Deposition and Cellular Uptake Behavior of Pathogen-Mimicking Nanovaccines in the First 48 Hours. Advanced healthcare materials

62. Huntimer LM, Ross KA, Darling RJ, Winterwood NE, Boggiatto P, et al. 2014. Polyanhydride nanovaccine platform enhances antigen-specific cytotoxic T cell responses. Technology 02:171-5

63. Ulery BD, Phanse Y, Sinha A, Wannemuehler MJ, Narasimhan B, Bellaire BH. 2009. Polymer chemistry influences monocytic uptake of polyanhydride nanospheres. Pharm Res 26:683-90

64. Bhattarai N, Gunn J, Zhang M. 2010. Chitosan-based hydrogels for controlled, localized drug delivery. Adv Drug Deliv Rev 62:83-99

65. Kang ML, Cho CS, Yoo HS. 2009. Application of chitosan microspheres for nasal delivery of vaccines. Biotechnology advances 27:857-65

66. Slutter B, Bal S, Keijzer C, Mallants R, Hagenaars N, et al. 2010. Nasal vaccination with N-trimethyl chitosan and PLGA based nanoparticles: nanoparticle characteristics determine quality and strength of the antibody response in mice against the encapsulated antigen. Vaccine 28:6282-91

67. Agnihotri SA, Mallikarjuna NN, Aminabhavi TM. 2004. Recent advances on chitosan- based micro- and nanoparticles in drug delivery. Journal of controlled release : official journal of the Controlled Release Society 100:5-28

68. van der Lubben IM, Verhoef JC, van Aelst AC, Borchard G, Junginger HE. 2001. Chitosan microparticles for oral vaccination: preparation, characterization and preliminary in vivo uptake studies in murine Peyer's patches. Biomaterials 22:687-94

69. Bernkop-Schnurch A, Dunnhaupt S. 2012. Chitosan-based drug delivery systems. European journal of pharmaceutics and biopharmaceutics : official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik e.V 81:463-9

70. Amidi M, Romeijn SG, Verhoef JC, Junginger HE, Bungener L, et al. 2007. N-trimethyl chitosan (TMC) nanoparticles loaded with influenza subunit antigen for intranasal vaccination: biological properties and immunogenicity in a mouse model. Vaccine 25:144-53

71. van der Lubben IM, Verhoef JC, Borchard G, Junginger HE. 2001. Chitosan and its derivatives in mucosal drug and vaccine delivery. European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences 14:201-7

72. Gupta NK, Tomar P, Sharma V, Dixit VK. 2011. Development and characterization of chitosan coated poly-(epsilon-caprolactone) nanoparticulate system for effective immunization against influenza. Vaccine 29:9026-37

73. Mahapatro A, Singh DK. 2011. Biodegradable nanoparticles are excellent vehicle for site directed in-vivo delivery of drugs and vaccines. J Nanobiotechnol 9

74. Balthasar S, Michaelis K, Dinauer N, von Briesen H, Kreuter J, Langer K. 2005. Preparation and characterisation of antibody modified gelatin nanoparticles as drug carrier system for uptake in lymphocytes. Biomaterials 26:2723-32

75. Wang J, Tauchi Y, Deguchi Y, Morimoto K, Tabata Y, Ikada Y. 2000. Positively charged gelatin microspheres as gastric mucoadhesive drug delivery system for eradication of H- pylori. Drug Deliv 7:237-43

76. Coester CJ, Langer K, Von Briesen H, Kreuter J. 2000. Gelatin nanoparticles by two step desolvation - a new preparation method, surface modifications and cell uptake. Journal of microencapsulation 17:187-93

77. Jahanshahi M, Babaei Z. 2008. Protein nanoparticle: A unique system as drug delivery vehicles. Afr J Biotechnol 7:4926-34

78. Coester C, Nayyar P, Samuel J. 2006. In vitro uptake of gelatin nanoparticles by murine dendritic cells and their intracellular localisation. European Journal of Pharmaceutics and Biopharmaceutics 62:306-14

79. Kaul G, Amiji M. 2002. Long-circulating poly(ethylene glycol)-modified gelatin nanoparticles for intracellular delivery. Pharm Res 19:1061-7

80. Leo E, Vandelli MA, Cameroni R, Forni F. 1997. Doxorubicin-loaded gelatin nanoparticles stabilized by glutaraldehyde: Involvement of the drug in the cross-linking process. International journal of pharmaceutics 155:75-82

81. Henriksen-Lacey M, Korsholm KS, Andersen P, Perrie Y, Christensen D. 2011. Liposomal vaccine delivery systems. Expert Opin Drug Del 8:505-19

82. Agger EM, Rosenkrands I, Hansen J, Brahimi K, Vandahl BS, et al. 2008. Cationic liposomes formulated with synthetic mycobacterial cordfactor (CAF01): a versatile adjuvant for vaccines with different immunological requirements. PloS one 3:e3116

83. Kaufmann SH. 2012. Tuberculosis vaccine development: strength lies in tenacity. Trends in immunology 33:373-9

84. Vermuri S, Rhodes CT. 1995. Preparation and Characterization of Liposomes as Therapeutic Delivery Systems: A Review. Pharmaceutica Acta Helvetiae:95-111

85. Torchilin V. 2012. Liposomes in Drug Delivery. Adv Del Sci Technol:289-328

86. Altin JG, Parish CR. 2006. Liposomal vaccines - targeting the delivery of antigen. Methods 40:39-52

87. Nakanishi T, Kunisawa J, Hayashi A, Tsutsumi Y, Kubo K, et al. 1999. Positively charged liposome functions as an efficient immunoadjuvant in inducing cell-mediated immune response to soluble proteins. Journal of controlled release : official journal of the Controlled Release Society 61:233-40

88. Henriksen-Lacey M, Bramwell VW, Christensen D, Agger EM, Andersen P, Perrie Y. 2010. Liposomes based on dimethyldioctadecylammonium promote a depot effect and enhance immunogenicity of soluble antigen. Journal of controlled release : official journal of the Controlled Release Society 142:180-6

89. Mallapragada SK, Narasimhan B. 2008. Immunomodulatory biomaterials. International journal of pharmaceutics 364:265-71

90. van Vliet SJ, Garcia-Vallejo JJ, van Kooyk Y. 2008. Dendritic cells and C-type lectin receptors: coupling innate to adaptive immune responses. Immunology and cell biology 86:580-7

91. Zamze S, Martinez-Pomares L, Jones H, Taylor PR, Stillion RJ, et al. 2002. Recognition of bacterial capsular polysaccharides and lipopolysaccharides by the macrophage mannose receptor. The Journal of biological chemistry 277:41613-23

92. Fiani ML, Beitz J, Turvy D, Blum JS, Stahl PD. 1998. Regulation of mannose receptor synthesis and turnover in mouse J774 macrophages. Journal of leukocyte biology 64:85-91

93. van de Veerdonk FL, Marijnissen RJ, Kullberg BJ, Koenen HJ, Cheng SC, et al. 2009. The macrophage mannose receptor induces IL-17 in response to Candida albicans. Cell host & microbe 5:329-40

94. Lee SJ, Evers S, Roeder D, Parlow AF, Risteli J, et al. 2002. Mannose receptor-mediated regulation of serum glycoprotein homeostasis. Science 295:1898-901

95. Gadjeva M, Takahashi K, Thiel S. 2004. Mannan-binding lectin--a soluble pattern recognition molecule. Molecular immunology 41:113-21

96. Ma YG, Cho MY, Zhao M, Park JW, Matsushita M, et al. 2004. Human mannose- binding lectin and L-ficolin function as specific pattern recognition proteins in the lectin activation pathway of complement. The Journal of biological chemistry 279:25307-12

97. Takahashi M, Iwaki D, Kanno K, Ishida Y, Xiong J, et al. 2008. Mannose-binding lectin (MBL)-associated serine protease (MASP)-1 contributes to activation of the lectin complement pathway. Journal of immunology 180:6132-8

98. Tacken PJ, de Vries IJ, Torensma R, Figdor CG. 2007. Dendritic-cell immunotherapy: from ex vivo loading to in vivo targeting. Nature reviews. Immunology 7:790-802

99. Mahnke K, Guo M, Lee S, Sepulveda H, Swain SL, et al. 2000. The dendritic cell receptor for endocytosis, DEC-205, can recycle and enhance antigen presentation via major histocompatibility complex class II-positive lysosomal compartments. The Journal of cell biology 151:673-84

100. Kato M, McDonald KJ, Khan S, Ross IL, Vuckovic S, et al. 2006. Expression of human DEC-205 (CD205) multilectin receptor on leukocytes. International immunology 18:857-69

101. Gliddon DR, Hope JC, Brooke GP, Howard CJ. 2004. DEC-205 expression on migrating dendritic cells in afferent lymph. Immunology 111:262-72

102. Birkholz K, Schwenkert M, Kellner C, Gross S, Fey G, et al. 2010. Targeting of DEC- 205 on human dendritic cells results in efficient MHC class II-restricted antigen presentation. Blood 116:2277-85

103. Geijtenbeek TB, van Vliet SJ, Engering A, t Hart BA, van Kooyk Y. 2004. Self- and nonself-recognition by C-type lectins on dendritic cells. Annual review of immunology 22:33- 54

104. Tabarani G, Reina JJ, Ebel C, Vives C, Lortat-Jacob H, et al. 2006. Mannose hyperbranched dendritic polymers interact with clustered organization of DC-SIGN and inhibit gp120 binding. FEBS letters 580:2402-8

105. Feinberg H, Guo Y, Mitchell DA, Drickamer K, Weis WI. 2005. Extended neck regions stabilize tetramers of the receptors DC-SIGN and DC-SIGNR. The Journal of biological chemistry 280:1327-35

106. Azad AK, Torrelles JB, Schlesinger LS. 2008. Mutation in the DC-SIGN cytoplasmic triacidic cluster motif markedly attenuates receptor activity for phagocytosis and endocytosis of mannose-containing ligands by human myeloid cells. Journal of leukocyte biology 84:1594-603

107. Reid DM, Montoya M, Taylor PR, Borrow P, Gordon S, et al. 2004. Expression of the beta-glucan receptor, Dectin-1, on murine leukocytes in situ correlates with its function in pathogen recognition and reveals potential roles in leukocyte interactions. Journal of leukocyte biology 76:86-94

108. Herre J, Marshall AS, Caron E, Edwards AD, Williams DL, et al. 2004. Dectin-1 uses novel mechanisms for yeast phagocytosis in macrophages. Blood 104:4038-45

109. Gringhuis SI, Wevers BA, Kaptein TM, van Capel TM, Theelen B, et al. 2011. Selective C-Rel activation via Malt1 controls anti-fungal T(H)-17 immunity by dectin-1 and dectin-2. PLoS pathogens 7:e1001259

110. O'Neill LA, Golenbock D, Bowie AG. 2013. The history of Toll-like receptors - redefining innate immunity. Nature reviews. Immunology 13:453-60

111. Baumann CL, Aspalter IM, Sharif O, Pichlmair A, Bluml S, et al. 2010. CD14 is a coreceptor of Toll-like receptors 7 and 9. The Journal of experimental medicine 207:2689-701

112. Bernasconi NL, Onai N, Lanzavecchia A. 2003. A role for Toll-like receptors in acquired immunity: up-regulation of TLR9 by BCR triggering in naive B cells and constitutive expression in memory B cells. Blood 101:4500-4

113. Komai-Koma M, Jones L, Ogg GS, Xu D, Liew FY. 2004. TLR2 is expressed on activated T cells as a costimulatory receptor. Proceedings of the National Academy of Sciences of the United States of America 101:3029-34

114. Nunez Miguel R, Wong J, Westoll JF, Brooks HJ, O'Neill LA, et al. 2007. A dimer of the Toll-like receptor 4 cytoplasmic domain provides a specific scaffold for the recruitment of signalling adaptor proteins. PloS one 2:e788

115. Lebeis SL, Bommarius B, Parkos CA, Sherman MA, Kalman D. 2007. TLR signaling mediated by MyD88 is required for a protective innate immune response by neutrophils to Citrobacter rodentium. Journal of immunology 179:566-77

116. Higgins SC, Jarnicki AG, Lavelle EC, Mills KH. 2006. TLR4 mediates vaccine-induced protective cellular immunity to Bordetella pertussis: role of IL-17-producing T cells. Journal of immunology 177:7980-9

117. Kurt-Jones EA, Popova L, Kwinn L, Haynes LM, Jones LP, et al. 2000. Pattern recognition receptors TLR4 and CD14 mediate response to respiratory syncytial virus. Nature immunology 1:398-401

118. Alderson MR, McGowan P, Baldridge JR, Probst P. 2006. TLR4 agonists as immunomodulatory agents. Journal of endotoxin research 12:313-9

119. Ohashi K, Burkart V, Flohe S, Kolb H. 2000. Cutting edge: heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex. Journal of immunology 164:558-61

120. Farhat K, Riekenberg S, Heine H, Debarry J, Lang R, et al. 2008. Heterodimerization of TLR2 with TLR1 or TLR6 expands the ligand spectrum but does not lead to differential signaling. Journal of leukocyte biology 83:692-701

121. Yoon SI, Kurnasov O, Natarajan V, Hong M, Gudkov AV, et al. 2012. Structural basis of TLR5-flagellin recognition and signaling. Science 335:859-64

122. Van Maele L, Carnoy C, Cayet D, Songhet P, Dumoutier L, et al. 2010. TLR5 signaling stimulates the innate production of IL-17 and IL-22 by CD3(neg)CD127+ immune cells in spleen and mucosa. Journal of immunology 185:1177-85

123. Stahl-Hennig C, Eisenblatter M, Jasny E, Rzehak T, Tenner-Racz K, et al. 2009. Synthetic double-stranded RNAs are adjuvants for the induction of T helper 1 and humoral immune responses to human papillomavirus in rhesus macaques. PLoS pathogens 5:e1000373

124. Liu L, Botos I, Wang Y, Leonard JN, Shiloach J, et al. 2008. Structural basis of toll-like receptor 3 signaling with double-stranded RNA. Science 320:379-81

125. Jelinek I, Leonard JN, Price GE, Brown KN, Meyer-Manlapat A, et al. 2011. TLR3- specific double-stranded RNA oligonucleotide adjuvants induce dendritic cell cross- presentation, CTL responses, and antiviral protection. Journal of immunology 186:2422-9

126. Chen K, Xiang Y, Yao X, Liu Y, Gong W, et al. 2011. The active contribution of Toll- like receptors to allergic airway inflammation. International immunopharmacology 11:1391-8

127. Kowalinski E, Lunardi T, McCarthy AA, Louber J, Brunel J, et al. 2011. Structural basis for the activation of innate immune pattern-recognition receptor RIG-I by viral RNA. Cell 147:423-35

128. Kato H, Takeuchi O, Sato S, Yoneyama M, Yamamoto M, et al. 2006. Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441:101-5

129. Satoh T, Kato H, Kumagai Y, Yoneyama M, Sato S, et al. 2010. LGP2 is a positive regulator of RIG-I- and MDA5-mediated antiviral responses. Proceedings of the National Academy of Sciences of the United States of America 107:1512-7

130. Binder M, Eberle F, Seitz S, Mucke N, Huber CM, et al. 2011. Molecular mechanism of signal perception and integration by the innate immune sensor retinoic acid-inducible gene-I (RIG-I). The Journal of biological chemistry 286:27278-87

131. Chiu YH, Macmillan JB, Chen ZJ. 2009. RNA polymerase III detects cytosolic DNA and induces type I interferons through the RIG-I pathway. Cell 138:576-91

132. Jiang X, Kinch LN, Brautigam CA, Chen X, Du F, et al. 2012. Ubiquitin-induced oligomerization of the RNA sensors RIG-I and MDA5 activates antiviral innate immune response. Immunity 36:959-73

133. Yoneyama M, Fujita T. 2009. RNA recognition and signal transduction by RIG-I-like receptors. Immunological reviews 227:54-65

134. Warren SE, Mao DP, Rodriguez AE, Miao EA, Aderem A. 2008. Multiple Nod-like receptors activate caspase 1 during Listeria monocytogenes infection. Journal of immunology 180:7558-64

135. Franchi L, Warner N, Viani K, Nunez G. 2009. Function of Nod-like receptors in microbial recognition and host defense. Immunological reviews 227:106-28

136. Petnicki-Ocwieja T, Hrncir T, Liu YJ, Biswas A, Hudcovic T, et al. 2009. Nod2 is required for the regulation of commensal microbiota in the intestine. Proceedings of the National Academy of Sciences of the United States of America 106:15813-8

137. Mo J, Boyle JP, Howard CB, Monie TP, Davis BK, Duncan JA. 2012. Pathogen sensing by nucleotide-binding oligomerization domain-containing protein 2 (NOD2) is mediated by direct binding to muramyl dipeptide and ATP. The Journal of biological chemistry 287:23057-67

138. Ferwerda G, Meyer-Wentrup F, Kullberg BJ, Netea MG, Adema GJ. 2008. Dectin-1 synergizes with TLR2 and TLR4 for cytokine production in human primary monocytes and macrophages. Cellular microbiology 10:2058-66

139. van der Meel R, Vehmeijer LJ, Kok RJ, Storm G, van Gaal EV. 2013. Ligand-targeted particulate nanomedicines undergoing clinical evaluation: current status. Adv Drug Deliv Rev 65:1284-98

140. Hogarth PM, Pietersz GA. 2012. Fc receptor-targeted therapies for the treatment of inflammation, cancer and beyond. Nature reviews. Drug discovery 11:311-31

141. Hong Y, Peng Y, Xiao H, Mi M, Munn D, He Y. 2012. Immunoglobulin Fc fragment tagging allows strong activation of endogenous CD4 T cells to reshape the tumor milieu and enhance the antitumor effect of lentivector immunization. Journal of immunology 188:4819- 27

142. Rawool DB, Bitsaktsis C, Li Y, Gosselin DR, Lin Y, et al. 2008. Utilization of Fc receptors as a mucosal vaccine strategy against an intracellular bacterium, Francisella tularensis. Journal of immunology 180:5548-57

143. Ejaz A, Ammann CG, Werner R, Huber G, Oberhauser V, et al. 2012. Targeting viral antigens to CD11c on dendritic cells induces retrovirus-specific T cell responses. PloS one 7:e45102

144. Castro FV, Tutt AL, White AL, Teeling JL, James S, et al. 2008. CD11c provides an effective immunotarget for the generation of both CD4 and CD8 T cell responses. European journal of immunology 38:2263-73

145. Daniels TR, Delgado T, Helguera G, Penichet ML. 2006. The transferrin receptor part II: targeted delivery of therapeutic agents into cancer cells. Clinical immunology 121:159-76

146. Normanno N, De Luca A, Bianco C, Strizzi L, Mancino M, et al. 2006. Epidermal growth factor receptor (EGFR) signaling in cancer. Gene 366:2-16

147. Carter PJ, Senter PD. 2008. Antibody-drug conjugates for cancer therapy. Cancer journal 14:154-69

148. Carter PJ. 2006. Potent antibody therapeutics by design. Nature reviews. Immunology 6:343-57

149. Reichert JM, Rosensweig CJ, Faden LB, Dewitz MC. 2005. Monoclonal antibody successes in the clinic. Nature biotechnology 23:1073-8

150. Palmer DH, Midgley RS, Mirza N, Torr EE, Ahmed F, et al. 2009. A phase II study of adoptive immunotherapy using dendritic cells pulsed with tumor lysate in patients with hepatocellular carcinoma. Hepatology 49:124-32

151. Raich-Regue D, Naranjo-Gomez M, Grau-Lopez L, Ramo C, Pujol-Borrell R, et al. 2012. Differential effects of monophosphoryl lipid A and cytokine cocktail as maturation stimuli of immunogenic and tolerogenic dendritic cells for immunotherapy. Vaccine 30:378-87

152. Fucikova J, Rozkova D, Ulcova H, Budinsky V, Sochorova K, et al. 2011. Poly I: C- activated dendritic cells that were generated in CellGro for use in cancer immunotherapy trials. Journal of translational medicine 9:223

153. Thomas-Kaskel AK, Zeiser R, Jochim R, Robbel C, Schultze-Seemann W, et al. 2006. Vaccination of advanced prostate cancer patients with PSCA and PSA peptide-loaded dendritic cells induces DTH responses that correlate with superior overall survival. International journal of cancer. Journal international du cancer 119:2428-34

154. Restifo NP, Dudley ME, Rosenberg SA. 2012. Adoptive immunotherapy for cancer: harnessing the T cell response. Nature reviews. Immunology 12:269-81

155. Adams EW, Ratner DM, Seeberger PH, Hacohen N. 2008. Carbohydrate-mediated targeting of antigen to dendritic cells leads to enhanced presentation of antigen to T cells. Chembiochem : a European journal of chemical biology 9:294-303

156. Bonifaz LC, Bonnyay DP, Charalambous A, Darguste DI, Fujii S, et al. 2004. In vivo targeting of antigens to maturing dendritic cells via the DEC-205 receptor improves T cell vaccination. The Journal of experimental medicine 199:815-24

157. Carrillo-Conde B, Song EH, Chavez-Santoscoy A, Phanse Y, Ramer-Tait AE, et al. 2011. Mannose-functionalized "pathogen-like" polyanhydride nanoparticles target C-type lectin receptors on dendritic cells. Molecular pharmaceutics 8:1877-86

158. Singh SK, Stephani J, Schaefer M, Kalay H, Garcia-Vallejo JJ, et al. 2009. Targeting glycan modified OVA to murine DC-SIGN transgenic dendritic cells enhances MHC class I and II presentation. Molecular immunology 47:164-74

159. Singh SK, Streng-Ouwehand I, Litjens M, Kalay H, Saeland E, van Kooyk Y. 2011. Tumour-associated glycan modifications of antigen enhance MGL2 dependent uptake and MHC class I restricted CD8 T cell responses. International journal of cancer. Journal international du cancer 128:1371-83

160. Kortylewski M, Swiderski P, Herrmann A, Wang L, Kowolik C, et al. 2009. In vivo delivery of siRNA to immune cells by conjugation to a TLR9 agonist enhances antitumor immune responses. Nature biotechnology 27:925-32

161. Higgins SC, Mills KH. 2010. TLR, NLR Agonists, and Other Immune Modulators as Infectious Disease Vaccine Adjuvants. Current infectious disease reports 12:4-12

162. Kufer TA, Sansonetti PJ. 2007. Sensing of bacteria: NOD a lonely job. Current opinion in microbiology 10:62-9

163. Vecchi S, Bufali S, Uno T, Wu T, Arcidiacono L, et al. 2014. Conjugation of a TLR7 agonist and antigen enhances protection in the S. pneumoniae murine infection model. European journal of pharmaceutics and biopharmaceutics : official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik e.V

164. Jia F, Liu X, Li L, Mallapragada S, Narasimhan B, Wang Q. 2013. Multifunctional nanoparticles for targeted delivery of immune activating and cancer therapeutic agents. Journal of controlled release : official journal of the Controlled Release Society 172:1020-34

165. Phanse Y, Carrillo-Conde BR, Ramer-Tait AE, Roychoudhury R, Pohl NL, et al. 2013. Functionalization of polyanhydride microparticles with di-mannose influences uptake by and intracellular fate within dendritic cells. Acta biomaterialia

166. Chavez-Santoscoy AV, Roychoudhury R, Pohl NL, Wannemuehler MJ, Narasimhan B, Ramer-Tait AE. 2012. Tailoring the immune response by targeting C-type lectin receptors on alveolar macrophages using "pathogen-like" amphiphilic polyanhydride nanoparticles. Biomaterials 33:4762-72

167. Tamayo I, Irache JM, Mansilla C, Ochoa-Reparaz J, Lasarte JJ, Gamazo C. 2010. Poly(Anhydride) Nanoparticles Act as Active Th1 Adjuvants through Toll-Like Receptor Exploitation. Clin Vaccine Immunol 17:1356-62

168. Raghuwanshi D, Mishra V, Das D, Kaur K, Suresh MR. 2012. Dendritic cell targeted chitosan nanoparticles for nasal DNA immunization against SARS CoV nucleocapsid protein. Molecular pharmaceutics 9:946-56

169. Demento SL, Bonafe N, Cui W, Kaech SM, Caplan MJ, et al. 2010. TLR9-targeted biodegradable nanoparticles as immunization vectors protect against West Nile encephalitis. Journal of immunology 185:2989-97

170. Bafica A, Scanga CA, Feng CG, Leifer C, Cheever A, Sher A. 2005. TLR9 regulates Th1 responses and cooperates with TLR2 in mediating optimal resistance to Mycobacterium tuberculosis. The Journal of experimental medicine 202:1715-24

171. Fahmy TM, Samstein RM, Harness CC, Mark Saltzman W. 2005. Surface modification of biodegradable polyesters with fatty acid conjugates for improved drug targeting. Biomaterials 26:5727-36

172. de Titta A, Ballester M, Julier Z, Nembrini C, Jeanbart L, et al. 2013. Nanoparticle conjugation of CpG enhances adjuvancy for cellular immunity and memory recall at low dose. Proceedings of the National Academy of Sciences of the United States of America 110:19902- 7

173. Pavot V, Rochereau N, Primard C, Genin C, Perouzel E, et al. 2013. Encapsulation of Nod1 and Nod2 receptor ligands into poly(lactic acid) nanoparticles potentiates their immune properties. Journal of controlled release : official journal of the Controlled Release Society 167:60-7

174. Kolli D, Velayutham TS, Casola A. 2013. Host-Viral Interactions: Role of Pattern Recognition Receptors (PRRs) in Human Pneumovirus Infections. Pathogens 2

175. Skountzou I, Martin Mdel P, Wang B, Ye L, Koutsonanos D, et al. 2010. Salmonella flagellins are potent adjuvants for intranasally administered whole inactivated influenza vaccine. Vaccine 28:4103-12

176. Turley CB, Rupp RE, Johnson C, Taylor DN, Wolfson J, et al. 2011. Safety and immunogenicity of a recombinant M2e-flagellin influenza vaccine (STF2.4xM2e) in healthy adults. Vaccine 29:5145-52

177. Treanor JJ, Taylor DN, Tussey L, Hay C, Nolan C, et al. 2010. Safety and immunogenicity of a recombinant hemagglutinin influenza-flagellin fusion vaccine (VAX125) in healthy young adults. Vaccine 28:8268-74

178. Coler RN, Baldwin SL, Shaverdian N, Bertholet S, Reed SJ, et al. 2010. A synthetic adjuvant to enhance and expand immune responses to influenza vaccines. PloS one 5:e13677

179. Ichinohe T, Watanabe I, Ito S, Fujii H, Moriyama M, et al. 2005. Synthetic double- stranded RNA poly(I:C) combined with mucosal vaccine protects against influenza virus infection. Journal of virology 79:2910-9

180. Carrat F, Flahault A. 2007. Influenza vaccine: the challenge of antigenic drift. Vaccine 25:6852-62

181. Taubenberger JK, Morens DM. 2008. The pathology of influenza virus infections. Annual review of pathology 3:499-522

182. Panda A, Qian F, Mohanty S, van Duin D, Newman FK, et al. 2010. Age-associated decrease in TLR function in primary human dendritic cells predicts influenza vaccine response. Journal of immunology 184:2518-27

183. Wong SS, Webby RJ. 2013. Traditional and new influenza vaccines. Clinical microbiology reviews 26:476-92

184. McElhaney JE. 2005. The unmet need in the elderly: designing new influenza vaccines for older adults. Vaccine 23 Suppl 1:S10-25

185. Acheson NH. 2011. Fundamentals of Molecular Virology. pp 500. John Wiley & Sons

186. Trifonov V, Khiabanian H, Rabadan R. 2009. Geographic dependence, surveillance, and origins of the 2009 influenza A (H1N1) virus. The New England journal of medicine 361:115- 9

187. Lemaitre M, Carrat F. 2010. Comparative age distribution of influenza morbidity and mortality during seasonal influenza epidemics and the 2009 H1N1 pandemic. BMC infectious diseases 10:162

188. Bardiya N, Bae JH. 2005. Influenza vaccines: recent advances in production technologies. Applied microbiology and biotechnology 67:299-305

189. Belshe RB. 2004. Current status of live attenuated influenza virus vaccine in the US. Virus research 103:177-85

190. Stevens J, Blixt O, Tumpey TM, Taubenberger JK, Paulson JC, Wilson IA. 2006. Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312:404-10

191. Hannoun C. 2013. The evolving history of influenza viruses and influenza vaccines. Expert review of vaccines 12:1085-94

192. Smith AE, Helenius A. 2004. How viruses enter animal cells. Science 304:237-42

193. Skehel JJ, Wiley DC. 2000. Receptor binding and membrane fusion in virus entry: the influenza hemagglutinin. Annual review of biochemistry 69:531-69

194. Swalley SE, Baker BM, Calder LJ, Harrison SC, Skehel JJ, Wiley DC. 2004. Full-length influenza hemagglutinin HA2 refolds into the trimeric low-pH-induced conformation. Biochemistry 43:5902-11

195. Hale BG, Randall RE, Ortin J, Jackson D. 2008. The multifunctional NS1 protein of influenza A viruses. The Journal of general virology 89:2359-76

196. Mibayashi M, Martinez-Sobrido L, Loo YM, Cardenas WB, Gale M, Jr., Garcia-Sastre A. 2007. Inhibition of retinoic acid-inducible gene I-mediated induction of beta interferon by the NS1 protein of influenza A virus. Journal of virology 81:514-24

197. Guo Z, Chen LM, Zeng H, Gomez JA, Plowden J, et al. 2007. NS1 protein of influenza A virus inhibits the function of intracytoplasmic pathogen sensor, RIG-I. American journal of respiratory cell and molecular biology 36:263-9

198. Sharma K, Tripathi S, Ranjan P, Kumar P, Garten R, et al. 2011. Influenza A virus nucleoprotein exploits Hsp40 to inhibit PKR activation. PloS one 6:e20215

199. Suzuki Y. 2005. Sialobiology of influenza: molecular mechanism of host range variation of influenza viruses. Biological & pharmaceutical bulletin 28:399-408

200. Hien TT, de Jong M, Farrar J. 2004. Avian influenza--a challenge to global health care structures. The New England journal of medicine 351:2363-5

201. Cox NJ, Subbarao K. 2000. Global epidemiology of influenza: past and present. Annual review of medicine 51:407-21

202. Harper SA, Bradley JS, Englund JA, File TM, Gravenstein S, et al. 2009. Seasonal influenza in adults and children--diagnosis, treatment, chemoprophylaxis, and institutional outbreak management: clinical practice guidelines of the Infectious Diseases Society of America. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 48:1003-32

203. Ichinohe T. 2010. Respective roles of TLR, RIG-I and NLRP3 in influenza virus infection and immunity: impact on vaccine design. Expert review of vaccines 9:1315-24

204. Subbarao K, Joseph T. 2007. Scientific barriers to developing vaccines against avian influenza viruses. Nature reviews. Immunology 7:267-78

205. Gaydos JC, Top FH, Jr., Hodder RA, Russell PK. 2006. Swine influenza a outbreak, Fort Dix, New Jersey, 1976. Emerging infectious diseases 12:23-8

206. Laver G, Garman E. 2002. Pandemic influenza: its origin and control. Microbes and infection / Institut Pasteur 4:1309-16

207. Itoh Y, Shinya K, Kiso M, Watanabe T, Sakoda Y, et al. 2009. In vitro and in vivo characterization of new swine-origin H1N1 influenza viruses. Nature 460:1021-5

208. Novel Swine-Origin Influenza AVIT, Dawood FS, Jain S, Finelli L, Shaw MW, et al. 2009. Emergence of a novel swine-origin influenza A (H1N1) virus in humans. The New England journal of medicine 360:2605-15

209. Woo PC, Tung ET, Chan KH, Lau CC, Lau SK, Yuen KY. 2010. Cytokine profiles induced by the novel swine-origin influenza A/H1N1 virus: implications for treatment strategies. The Journal of infectious diseases 201:346-53

210. Pollack A. 2009. Benefit and Doubt in Vaccine Additive.

211. Ulmer JB, Valley U, Rappuoli R. 2006. Vaccine manufacturing: challenges and solutions. Nature biotechnology 24:1377-83

212. Wanzeck K, Boyd KL, McCullers JA. 2011. Glycan shielding of the influenza virus hemagglutinin contributes to immunopathology in mice. American journal of respiratory and critical care medicine 183:767-73

213. Hutter J, Rodig JV, Hoper D, Seeberger PH, Reichl U, et al. 2013. Toward animal cell culture-based influenza vaccine design: viral hemagglutinin N-glycosylation markedly impacts immunogenicity. Journal of immunology 190:220-30

214. Couch RB, Winokur P, Brady R, Belshe R, Chen WH, et al. 2007. Safety and immunogenicity of a high dosage trivalent influenza vaccine among elderly subjects. Vaccine 25:7656-63

215. Cox RJ, Brokstad KA, Ogra P. 2004. Influenza virus: immunity and vaccination strategies. Comparison of the immune response to inactivated and live, attenuated influenza vaccines. Scandinavian journal of immunology 59:1-15

216. Belshe RB, Edwards KM, Vesikari T, Black SV, Walker RE, et al. 2007. Live attenuated versus inactivated influenza vaccine in infants and young children. The New England journal of medicine 356:685-96

217. Rhorer J, Ambrose CS, Dickinson S, Hamilton H, Oleka NA, et al. 2009. Efficacy of live attenuated influenza vaccine in children: A meta-analysis of nine randomized clinical trials. Vaccine 27:1101-10

218. Cao RG, Suarez NM, Obermoser G, Lopez SMC, Flano E, et al. 2014. Differences in Antibody Responses Between Trivalent Inactivated Influenza Vaccine and Live Attenuated Influenza Vaccine Correlate With the Kinetics and Magnitude of Interferon Signaling in Children. Journal of Infectious Diseases 210:224-33

219. Ambrose CS, Wu X, Knuf M, Wutzler P. 2012. The efficacy of intranasal live attenuated influenza vaccine in children 2 through 17 years of age: a meta-analysis of 8 randomized controlled studies. Vaccine 30:886-92

220. Monto AS, Ohmit SE, Petrie JG, Johnson E, Truscon R, et al. 2009. Comparative Efficacy of Inactivated and Live Attenuated Influenza Vaccines. New Engl J Med 361:1260-7

221. Ambrose CS, Levin MJ, Belshe RB. 2011. The relative efficacy of trivalent live attenuated and inactivated influenza vaccines in children and adults. Influenza and other respiratory viruses 5:67-75

222. Halasa N, Englund JA, Nachman S, Weinberg GA, Huber VC, et al. 2011. Safety of live attenuated influenza vaccine in mild to moderately immunocompromised children with cancer. Vaccine 29:4110-5

223. Ruggiero G, Utili R. 1992. [Usefulness and limitations of vaccination against influenza]. Recenti progressi in medicina 83:337-40

224. Kost TA, Condreay JP, Jarvis DL. 2005. Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nature biotechnology 23:567-75

225. van Oers MM. 2011. Opportunities and challenges for the baculovirus expression system. J Invertebr Pathol 107:S3-S15

226. Lin SC, Jan JT, Dionne B, Butler M, Huang MH, et al. 2013. Different immunity elicited by recombinant H5N1 hemagglutinin proteins containing pauci-mannose, high-mannose, or complex type N-glycans. PloS one 8:e66719

227. Wang K, Holtz KM, Anderson K, Chubet R, Mahmoud W, Cox MM. 2006. Expression and purification of an influenza hemagglutinin--one step closer to a recombinant protein- based influenza vaccine. Vaccine 24:2176-85

228. Cox MM, Patriarca PA, Treanor J. 2008. FluBlok, a recombinant hemagglutinin influenza vaccine. Influenza and other respiratory viruses 2:211-9

229. Cox MM, Hollister JR. 2009. FluBlok, a next generation influenza vaccine manufactured in insect cells. Biologicals : journal of the International Association of Biological Standardization 37:182-9

230. Chichester JA, Haaheim LR, Yusibov V. 2009. Using plant cells as influenza vaccine substrates. Expert review of vaccines 8:493-8

231. Shoji Y, Chichester JA, Bi H, Musiychuk K, de la Rosa P, et al. 2008. Plant-expressed HA as a seasonal influenza vaccine candidate. Vaccine 26:2930-4

232. Mett V, Musiychuk K, Bi H, Farrance CE, Horsey A, et al. 2008. A plant-produced influenza subunit vaccine protects ferrets against virus challenge. Influenza and other respiratory viruses 2:33-40

233. Sokolova V, Knuschke T, Kovtun A, Buer J, Epple M, Westendorf AM. 2010. The use of calcium phosphate nanoparticles encapsulating Toll-like receptor ligands and the antigen hemagglutinin to induce dendritic cell maturation and T cell activation. Biomaterials 31:5627- 33

234. Galloway AL, Murphy A, DeSimone JM, Di J, Herrmann JP, et al. 2013. Development of a nanoparticle-based influenza vaccine using the PRINT (R) technology. Nanomed- Nanotechnol 9:523-31

235. Garinot M, Fievez V, Pourcelle V, Stoffelbach F, des Rieux A, et al. 2007. PEGylated PLGA-based nanoparticles targeting M cells for oral vaccination. Journal of controlled release : official journal of the Controlled Release Society 120:195-204

236. Tritto E, Mosca F, De Gregorio E. 2009. Mechanism of action of licensed vaccine adjuvants. Vaccine 27:3331-4

237. Okamoto S, Yoshii H, Akagi T, Akashi M, Ishikawa T, et al. 2007. Influenza hemagglutinin vaccine with poly(gamma-glutamic acid) nanoparticles enhances the protection against influenza virus infection through both humoral and cell-mediated immunity. Vaccine 25:8270-8

238. Wack A, Baudner BC, Hilbert AK, Manini I, Nuti S, et al. 2008. Combination adjuvants for the induction of potent, long-lasting antibody and T-cell responses to influenza vaccine in mice. Vaccine 26:552-61

239. Frey SE, Reyes MR, Reynales H, Bermal NN, Nicolay U, et al. 2014. Comparison of the safety and immunogenicity of an MF59((R))-adjuvanted with a non-adjuvanted seasonal influenza vaccine in elderly subjects. Vaccine 32:5027-34

240. Calabro S, Tortoli M, Baudner BC, Pacitto A, Cortese M, et al. 2011. Vaccine adjuvants alum and MF59 induce rapid recruitment of neutrophils and monocytes that participate in antigen transport to draining lymph nodes. Vaccine 29:1812-23

241. O'Hagan DT, Ott GS, De Gregorio E, Seubert A. 2012. The mechanism of action of MF59 - an innately attractive adjuvant formulation. Vaccine 30:4341-8

242. Roman F, Vaman T, Gerlach B, Markendorf A, Gillard P, Devaster JM. 2010. Immunogenicity and safety in adults of one dose of influenza A H1N1v 2009 vaccine formulated with and without AS03A-adjuvant: preliminary report of an observer-blind, randomised trial. Vaccine 28:1740-5

243. Roman F, Clement F, Dewe W, Walravens K, Maes C, et al. 2011. Effect on cellular and humoral immune responses of the AS03 adjuvant system in an A/H1N1/2009 influenza virus vaccine administered to adults during two randomized controlled trials. Clin Vaccine Immunol 18:835-43

244. Daniell H, Singh ND, Mason H, Streatfield SJ. 2009. Plant-made vaccine antigens and biopharmaceuticals. Trends in plant science 14:669-79

245. Pushko P, Tumpey TM, Bu F, Knell J, Robinson R, Smith G. 2005. Influenza virus-like particles comprised of the HA, NA, and M1 proteins of H9N2 influenza virus induce protective immune responses in BALB/c mice. Vaccine 23:5751-9

246. Bright RA, Carter DM, Daniluk S, Toapanta FR, Ahmad A, et al. 2007. Influenza virus- like particles elicit broader immune responses than whole virion inactivated influenza virus or recombinant hemagglutinin. Vaccine 25:3871-8

2.6 Figures

Figure 2.1 Hydrolysis and condensation of polyesters (top) and polyanhydrides (bottom).

Figure 2.2 Drug release from bulk eroding polymers such as, polyesters (top) and drug released from surface eroding polymers such as polyanhydrides (bottom) when incubated in an aqueous solution over an extended period of time.

Figure 2.3 Methods of particle fabrication. A) Spray drying begins when the solvent and its dissolved polymer are atomized in the presence of a stream of warm air. Particles are formed in the drying chamber as the solvent evaporates and a cyclone is used to collect the particles. The air is filtered as it exits. B) In nanoprecipitation the polymer is dissolved in a solvent while the protein to be encapsulated is dispersed in this solvent. Particle formation occurs when exposing this solution to an anti-solvent and these particles are collected through vacuum filtration.

Figure 2.4 Representative image of a liposome. It consists of a lipid bilayer commonly composed of phosphatidylcholine, phosphatidylethanolamine, or phosphatidylserine. The inner core is aqueous which is good for encapsulating water soluble proteins or nucleic acids.

Figure 2.5 Expression of pattern recognition receptors on an immature antigen presenting cell

(APC). CLRs are surface receptors that have great diversity in structure and signaling. TLRs dimerize upon binding to their ligand in a tail-to-tail type organization and then signal through

TRAM/TRIF or MAL/MYD88. NF-kB and IRF3 translocate to the nucleus and play a major role

in the innate immune systems response to infection. Most TLRs are surface receptors but there are several that are expressed in endosomes. RIG-I from the RLR family and NOD1 as well as

NOD2 from the NLR family are cytosolic PRRs that exist in a folded conformation until their

PAMP binds. ATP dependent unfolding occurs followed by oligomerization allowing the

CARDs to signal downstream.

Figure 2.6 Representative illustrations of the large diversity of mechanisms that can be utilized for active targeting of antigen or drugs. Non-cell based active targeting strategies can fall under three categories including functionalized antibodies, functionalized antigen, or functionalized materials. There is considerable overlap between all three strategies.

Figure 2.7 Structural depiction of an influenza virion. It has 8 strands of RNA that are covered in nucleocapsid proteins and are also bound to the RNA polymerase complex. The surface of the virion is mostly covered in hemagglutinin (HA) and neuraminidase (NA). HA is a trimer with each component having two subunits, HA1 and HA2. HA binds to sialic acid domains through its globular head on the HA1 subunit. The fusion domain is expressed on HA2 which remains mostly hidden until binding to sialic acid domains inducing a conformational change.

2.7 Tables Table 2.1 Common classes of biodegradable polymers used in drug delivery.

Table 2.2 Pattern recognition receptors, their ligands, and their expression on antigen presenting cells.

PRR Receptor Ligand Cellular Expression Location Macrophage Mannose Low order Mannoses Immature APCs, tissue macrophages Surface Receptor Dendritic Cells, Langerhans Cells, DEC-205 Unknown Surface B Cells, NK cells CLRs DC-SIGN Higher order Mannoses, Mac-1 Immature APCs, tissue macrophages Surface Dendritic Cells, Macrophages, Dectin-1 β-glucans Surface Neutrophils Macrophage Galactose Terminal galactoses, N-acetylgalactosamine Immature APCs Surface Lectin Lipopolysaccharides, Monophosphoryl Monocytes, Myeloid Dendritic Cells, TLR4 Surface Lipid A, Glucopyranpsyl Lipid A M cells Monocytes, Myeloid Dendritic Cells, TLR2-TLR1 Complex Pam CSK , peptidoglycan, zymosan Surface 2 4 T cells

Monocytes, Myeloid Dendritic Cells, TLR2-TLR6 Complex Pam CSK , peptidoglycan, zymosan Surface 2 4 T cells TLRs TLR5 Flagellin Epithelial Cells Surface TLR3 Poly(IC), Poly(ICLC), dsRNA Dendritic Cells Endosome TLR9 CpG DNA Plasmacytoid Dendritic Cells, B cells Endosome TLR7 ssRNA, Imiquimod Plasmacytoid Dendritic Cells Endosome TLR8 ssRNA Circulating Monocytes Endosome Macrophages, Dendritic cells, TLR11 Flagellin, Profilin Endosome Liver and Kidney Epithelial Cells RIG-I short ssRNA, dsRNA, poly(da:dT) Epithelial Cells, Mast Cells Cytosolic RLRs MDA5 long dsRNA, Poly(IC) Epithelial Cells Cytosolic NOD1 meso-diaminopemilic acid (DAP) Epithelial Cells, Paneth Cells Cytosolic NOD2 muramyl dipeptide (MDP) Monocytes Cytosolic NLRs APAF-1 cytochrome c Epithelial Cells, Paneth Cells Cytosolic Ipaf Flagellin Epithelial Cells, Paneth Cells Cytosolic NLRP3 ssRNA, extracellular ATP Epithelial Cells, Paneth Cells Cytosolic 86

Table 2.3 Chemical techniques used to functionalize drug delivery vehicles with ligands.

1) Intermediates are 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N- hydroxysulfosuccinimide (Sulfo-NHS). 2) Heterobifunctional linker that is 4-(4-N-maleimidophenyl)butyric acid hydrazide (MBPH). 4) Heterobifunctional linker that is succinimidyle-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC) 5) Homobifunctional linker that is glutaraldehyde 6) Intermediate that is 4-Toluenesulfonyl chloride (TsCl)

Table 2.4 The function of proteins expressed on an influenza virion.

Protein/Complex Abbreviation Location Function Surface of Lipid Responsible for membrane viral Hemagglutinin HA Envelope and cellular membrane fusion Surface of Lipid Cleaves virion from host cell receptor Neuraminidase NA Envelope to allow viral propagation Helps regulate virus uncoating by Ion Channel Transmembrane M2 allowing ions in Protein Protein or out of the lipid envelope Inner Membrane Interacts with nucleocapsids to help Matrix Protein M1 Protein assembly Nonstructural Within Lipid NS1 Suppresses host immune system Protein 1 Envelope Nonstructural Within Lipid Helps with viral RNA replication that NS2 Protein 2 Envelope occurs in the nucleus Nucleocapsid Bound to viral RNA and influences NP Bound to RNA Protein viral RNA synthesis RNA Polymerase Directly synthesizes messenger RNAs PA, PB1, PB2 Bound to RNA Complex andinvolved in genome replication

CHAPTER 3: RESEARCH OBJECTIVE AND THESIS ORGANIZATION

3.1 Research Objectives The overall goal of this research is to synthesize targeted nanoparticle-based vaccines against influenza virus. Even though there are a large number and a wide variety of influenza vaccines manufactured for human use every year, the virus continues to pose a significant public health problem. First, it has the ability to rapidly change its antigenic structure thereby avoiding the immune system, and secondly it has a large presence in animal reservoirs making zoonosis a constant threat. Current vaccines are limited, because they have different levels of efficacy across age groups and are notably weaker in the elderly. The intramuscular vaccines induce poor mucosal immune responses, and generally, they need to be stored cold. Additionally, the manufacturing process requires several months of lead time, putting significant strain on the ability to match the vaccine to circulating strains. A new generation of rationally designed influenza vaccines is necessary to combat these threats. In this regard, the use of nanovaccine technologies in influenza vaccine development shows great promise. Nanovaccine technologies can adjuvant the immune response and provide dose-sparing. This could help alleviate the burden on manufacturers because lower amounts of antigen will be required, while simultaneously increasing the efficacy, especially in at-risk populations. Nanovaccines have also been successfully administered by different routes, including intranasally, which is important for the induction of mucosal immune responses. This could improve patient compliance for influenza vaccines. The rational design of a safe influenza A vaccine utilizing the polyanhydride nanovaccine platform that induces long lasting protective immunity is described in this thesis.

The study has been organized into several specific goals (SGs) as shown below: 89

SG1: Study the interactions between carbohydrate-functionalized nanoparticles and serum proteins and their effects on dendritic cell activation;

SG2: Investigate the safety and biocompatibility of carbohydrate-functionalized nanoparticles in vivo;

SG3: Evaluate the ability of carbohydrate-functionalized nanoparticle-based formulations to induce humoral and cell-mediated immunity; and

SG4: Test the efficacy of lead candidate targeted nanovaccine formulations using the hemagglutinin antigen.

SG1 describes in vitro studies that focused on determining the role of the interactions between polymer chemistry, carbohydrate functionalization, and serum protein adsorption on dendritic cell activation. These studies identified nanoparticle formulations that are efficiently internalized by dendritic cells and that induced dendritic cell maturation. In SG2, the safety and biocompatibility of these carbohydrate-functionalized nanoparticles were evaluated by analyzing the histopathology of the liver, kidneys, and lungs of mice administered with these nanoparticle treatments. The cytokine and chemokine secretion in the bronchoalveolar lavage fluid as well as biomarkers in the blood and urine were analyzed for histopathological changes. In SG3, the ability of the carbohydrate-functionalized to induce potent cellular and humoral immune responses was evaluated using a model antigen, ovalbumin. In SG4, the results from the studies described in SG1, SG2, and SG3 were utilized to design efficacious nanoparticle-based formulations containing the influenza hemagglutinin antigen based on immune response evaluations as well as live viral challenge.

3.2 Thesis Organization The rationale, methods, results, and conclusions of the four specific goals are discussed in the following three chapters. Chapter 4 focuses on SG1, which covers the in vitro cell culture and cellular stimulation with the carbohydrate-modified nanoparticles. Chapter 5 outlines the completion of SG2 and discusses the safety and biocompatibility of the carbohydrate-modified nanoparticles. The ability of carbohydrate-modified nanoparticles to adjuvant the in vivo immune response shown by SG3 was evaluated in Chapter 6. Chapter 7 describes the use of a novel robot to synthesize combinatorial libraries of nanoparticles. This robot can surface functionalize nanoparticles which will aid in the screening and selection of functional groups that can target receptors on cells. Chapter 8 summarizes SG4 and evaluating where mice immunized with hemagglutinin loaded nanoparticles can withstand a live viral challenge. Chapter 9 highlights ongoing and future work and summarizes the results of this thesis.

CHAPTER 4: NANOPARTICLE CHEMISTRY AND

FUNCTIONALIZATION DIFFERENTIALLY REGULATES DENDRITIC

CELL-NANOPARTICLE INTERACTIONS AND TRIGGERS DENDRITIC

CELL MATURATION

Jonathan T. Goodman1, Julia E. Vela Ramirez1, Paola M. Boggiatto2, Rajarshi Roychoudhury3,

Nicola L. B. Pohl3, Michael J. Wannemuehler2, Balaji Narasimhan1

Reprinted from Particle and Particle Systems Characterization 31:1269–1280, 2014

1 Department of Chemical & Biological Engineering, Iowa State University, Ames, Iowa 50011 2 Department of Veterinary Microbiology & Preventive Medicine, Iowa State University, Ames, Iowa 50011 3 Department of Chemistry, Indiana University, Bloomington, Indiana 47405 92

4.1 Abstract

Biodegradable polyanhydride nanoparticles have been used as vaccine adjuvants and delivery vehicles. Upon parenteral administration, the surface characteristics of these nanoparticles may be modified by serum protein adsorption, likely affecting how the particles interact with antigen presenting cells. These studies investigated the role of the differential adsorption of serum proteins onto carbohydrate-functionalized and non-functionalized polyanhydride nanoparticles of different chemistries on interactions with dendritic cells (DCs).

Di-mannose (a carboxymethyl-modified disaccharide of 1,2-α-linked-D-mannopyranoside) or glycolic acid was covalently linked to nanoparticles synthesized using copolymers of 1,6-bis(p- carboxyphenoxy)hexane (CPH) and sebacic acid (SA) and 1,8-bis(p-carboxyphenoxy)-3,6- dioxaoctane (CPTEG) and CPH. Serum protein adsorption to CPH:SA nanoparticles enhanced particle internalization, but did not enhance uptake of CPTEG:CPH nanoparticles. When DCs were cultured with nanoparticles, the surface expression of MHC II, CD86, and CD206 was enhanced and mediated by both polymer chemistry and functionalization. Cytokine secretion by

DCs was differentially modulated by both polymer chemistry and functionalization. The amount and profile of serum proteins adsorbed indicated that polymer hydrophobicity and functionalization enhanced protein adsorption. Overall polymer chemistry, functionalization, and serum protein adsorption influenced nanoparticle interactions with DCs and elucidating the complex relationships between these parameters and dendritic cell maturation will enable rational design of targeted vaccines and immunotherapies. 93

4.2 Introduction

Vaccine delivery vehicles based on biodegradable polymeric nanoparticles present several advantages over traditional vaccines, including stabilization of encapsulated antigens, tailored erosion profiles leading to sustained antigen release, and enhanced adjuvant activity (1;

2). In particular, biodegradable polyanhydride nanoparticles have demonstrated the ability to induce immunomodulatory activity, be internalized effectively by antigen presenting cells, and provide dose-sparing capabilities (3-5).

Dendritic cells (DCs) have the ability to internalize, process, and present antigen to naïve

T cells, establishing them as a powerful link between the innate and adaptive immune responses

(6; 7). Dendritic cells use pattern recognition receptors (PRRs) to bind microbial associated molecular patterns (MAMPs) that in turn result in cellular activation. Examples of PRRs include

Toll-like receptors (TLRs), NOD-like receptors (NLRs), and C-type lectin receptors (CLRs) (6).

Specifically, CLRs comprise a diverse group of over 60 receptors that bind to carbohydrate moieties using highly conserved carbohydrate recognition domains (8). Many of them function by internalizing antigen for loading onto major histocompatibility complex (MHC) class I and II molecules while others function in tandem with other PRRs to influence intracellular signaling

(9).

An oft-used strategy in vaccine development is to use specific ligands to target PRRs, such as, CLRs, in order to impart potent adjuvant properties (10; 11). This “pathogen- mimicking” strategy allows for more efficient processing and presentation of antigen in the context of MHC I and II and interaction with T cells. This immune enhancing activity has been achieved previously by using delivery vehicles whose surfaces have been modified by attaching anti-CLR antibodies or carbohydrate moieties. While antibody-mediated targeting of CLRs

offers a high degree of specificity, a major limitation of this approach is that antibodies elicit immune responses that neutralize their effectiveness because of their inherent immunogenicity

(12). In contrast, carbohydrate ligands are generally non-immunogenic and can be more readily synthesized in pure form (12). Previous studies have demonstrated the effectiveness of functionalizing drug and/or vaccine delivery vehicles with carbohydrates (13-16). For example, di-mannose is present on the surface of Mycobacterium tuberculosis (and other pathogens) and is known to target a CLR, the macrophage mannose receptor (17). The macrophage mannose receptor, also known as CD206, is commonly expressed on immature DCs and macrophages and is known to bind to lower order mannose moieties and mediate internalization of microorganisms expressing mannosylated glycoproteins (18; 19). This receptor-mediated internalization mechanism can be exploited to enhance the ensuing immune response.

Recent work from our laboratory has shown that differential protein adsorption onto polyanhydride particles is mediated by polymer chemistry, which in turn influences the in vitro activation (e.g., phagocytosis, cell surface marker expression, cytokine secretion) of DCs (20;

21). It has also been shown that copolymers based on sebacic acid (SA) and 1,6-bis(p- carboxyphenoxy)hexane (CPH) and 1,8-bis(p-carboxyphenoxy)-3,6-dioxaoctane (CPTEG) and

CPH (Figure 4.1A) have different characteristics that could potentially affect the immune response (3; 13; 21; 22). Contact angle and degradation kinetics data have demonstrated that the

20:80 CPTEG:CPH chemistry is significantly more hydrophobic than the 20:80 CPH:SA chemistry (23; 24). These two formulations have been shown to be potent adjuvants, based on their ability to enhance both humoral and cell-mediated immunity (3; 4; 25).

However, the role of nanoparticle surface modification following serum protein adsorption on DC activation is yet unclear. It is known that serum proteins will adsorb onto the

surfaces of parenterally administered nanoparticles and affect how nanoparticles interact with antigen presenting cells and the nature of the resulting immune response. For example, the binding of complement proteins or immunoglobulins could provide important ligands to enhance antigen presenting cell-mediated phagocytosis of the nanoparticles and the subsequent processing of the encapsulated antigen. The purpose of this study was to investigate the role of polymer chemistry and carbohydrate functionalization on interactions with DCs. We hypothesized that nanoparticle surface functionalization would enhance the interactions between polyanhydride nanoparticles and DCs resulting in differential modulation of DC maturation and polyfunctionality as compared to non-functionalized particles. The current studies demonstrate that polymer chemistry-dependent functionalization affect DC biology by differentially regulating DC surface marker expression and cytokine production. This work provides insights into the in vivo performance of functionalized nanoparticles and demonstrates that these particles serve as more than vaccine-delivery vehicles as they directly interact with and modulate DC responses.

4.3. Materials and Methods

4.3.1 Materials

The chemicals needed for monomer synthesis included 1,6-dibromohexane, triethylene glycol, 4-p-hydroxybenzoic acid, and 1-methyl-2-pyrrolidinone which were purchased from

Sigma-Aldrich (St. Louis, MO). 4-p-fluorobenzonitrile was obtained from Apollo Scientific

(Cheshire, UK), while toluene, sulfuric acid, acetonitrile, dimethyl formamide, acetic anhydride, methylene chloride, pentane, and potassium carbonate were obtained from Fisher Scientific

(Fairlawn, NJ). Chemicals used for polymerization and nanoparticle synthesis included petroleum ether, pentane, acetic anhydride, chloroform, and methylene chloride, and they were

also purchased from Fisher Scientific. 1H NMR analysis was conducted using deuterated chloroform from Cambridge Isotope Laboratories (Andover, MA). Cadmium selenide quantum dots (QDs) emitting at 630 nm were kindly synthesized by Dr. Yanjie Zhang of Iowa State

University. Materials for nanoparticle functionalization included N-hydroxysuccinimide (NHS) from Fisher Scientific, ethylenediamine from Sigma Aldrich, 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (EDC) and glycolic acid from Acros Organics (Geel,

Belgium).

Materials for 2-D gel electrophoresis included 4-20% Tris-HCl linear gradient polyacrylamide gels, 7 cm non-linear pH 3-10 immobilized pH gradient (IPG) strips, Flamingo gel stain, and Precision Plus protein standards, all of which were obtained from BioRad

Laboratories (Richmond, CA). Rehydration buffer and sodium dodecyl sulfate equilibration buffer was formulated from urea, Bromophenol Blue, tris-Cl, glycerol, and sodium dodecyl sulfate was received from Fisher Scientific and CHAPS detergent was purchased from BioRad

Laboratories.

The DC culture medium was composed of β-mercaptoethanol (Sigma-Aldrich), granulocyte-macrophage colony-stimulating factor (GM-CSF) (PeproTech, Rocky Hill, NJ), heat inactivated fetal bovine serum (Atlanta Biologicals, Atlanta, GA), and RPMI 1640, penicillin/streptomycin, and L-glutamine (MediaTech, Herndon, VA). Escherichia coli lipopolysaccharide (LPS) 026:B6 was purchased from Sigma Aldrich.

Materials used in flow cytometry included BD stabilizing fixative solution from BD

Bioscience (San Jose, CA), while FITC anti-mouse CD11c (clone N418), antigen presenting cell anti-mouse CD206 (clone C068C2), PE-Cy7 anti-mouse CD86 (clone GL-1), FITC Armenian hamster IgG isotype control (clone HTK888), antigen presenting cell rat IgG2aκ isotype control

(clone RTK2758), PE-Cy7 rat IgG2aκ isotype control (clone RTK2758) were purchased from

Biolegend (San Diego, CA). V500 rat anti-mouse I-A/I-E (clone m5/114.15.2) and V500 IG2bκ isotype control (clone A95-1) were obtained from BD Bioscience.

4.3.2 Monomer and Polymer Synthesis

Sebacic acid (SA) was purchased from Sigma Aldrich, while CPH and CPTEG diacids were synthesized as described previously (23; 26; 27). Techniques described by Shen et al. and

Conix were used to synthesize SA and CPH prepolymers and a melt polycondensation reaction was carried out to synthesize 20:80 CPH:SA and 20:80 CPTEG:CPH copolymers (23; 28; 29).

Purity, chemical structure, and molecular weight of the polymers were determined using 1H

NMR spectroscopy on a Varian VXR 300 MHz spectrometer. The 20:80 CPH:SA copolymer had a molecular weight of 13.5 kDa, while the 20:80 CPTEG:CPH copolymer had a molecular weight of 7.5 kDa, which is consistent with previous work (5; 22). The actual molar ratios were determined to be 22:78 CPH:SA and 21:79 CPTEG:CPH using 1H NMR analysis.

4.3.3 Nanoparticle Synthesis:

Polyanhydride nanoparticles were synthesized using an anti-solvent nanoencapsulation technique (5). A 20 mg/mL polymer solution in methylene chloride was poured into a pentane bath so that the ratio of solvent to anti-solvent was 1:250. 20:80 CPH:SA nanoparticles were synthesized at room temperature, whereas 20:80 CPTEG:CPH nanoparticles were synthesized in a 4 °C cold room with the pentane bath held at -40 °C. For the QD-loaded particles, the synthesis procedures were identical to that of the blank nanoparticles except for the addition of 1% (w/w) of QDs into the methylene chloride to form QD-loaded nanoparticles. The particles were collected by vacuum filtration. Particle morphology was examined by a FEI Quanta 250 scanning electron microscope (SEM) (Kyoto, Japan) after being coated with 2 nm of iridium

using a Quorum Q150TS sputter coater (Lewes, UK). Nanoparticle size distribution was characterized using ImageJ image analysis software (National Institutes of Health, Bethesda,

MD) with an average of 200 nanoparticles per image.

4.3.4 Nanoparticle Functionalization

The particles were functionalized with either di-mannose or a glycolic acid linker, which links the di-mannose to the ethylenediamine (Figure 4.1C). Fluorous solid-phase extraction- based iterative synthesis was performed to obtain α-1,2-linked-di-mannopyranoside with a fluorous allyl group (30). The fluorous tag was cleaved by ozonolysis, while oxidation with

Jones Reagent produced a carboxylic acid group for coupling to nanoparticles. Birch reduction was used to yield a deprotected carboxymethyl-modified disaccharide of 1,2-α-linked-D- mannopyranoside (17; 30). A previously optimized two-step amide coupling reaction covalently linked the di-mannose to the nanoparticles (14; 15). The first reaction involved incubating the nanoparticles with 10 mole equivalents of EDC, 12 mole equivalents of NHS, and 10 mole equivalents of ethylenediamine for one hour on a tube rotor at 4 °C. The nanoparticles were then washed with nanopure water and were collected by centrifugation at 10,000 rcf and 10 min.

After being sonicated briefly at 40 Hz, the nanoparticles were incubated with 10 mole equivalents of EDC, 12 mole equivalents of NHS, and 10 mole equivalents of di-mannose for one hour at 4 °C allowing the di-mannose to be attached to the free amine of the ethylenediamine on the surface of the nanoparticles.

4.3.5 Characterization of Functionalized Nanoparticles

Quasi-elastic light scattering (QELS) using a Zetasizer Nano (Worchester, UK) was employed to determine the zeta potential (ζ) of the nanoparticles. A sonicated solution of 100

μg/mL of nanoparticles was suspended in cold water before measurement. The sugar density on

the surface of the nanoparticles was quantified using a phenol-sulfuric acid assay performed in flat bottom 96-well plates (31). A reaction time of 30 min at 95 °C was used for the sugars on the nanoparticle surface to be reduced inducing a color change in the solution. The absorbance of each well was measured at a wavelength of 490 nm using a SpectraMax M3 spectrophotometer

(Molecular Devices, Sunnyvale, CA). This was compared against soluble di-mannose standards, while non-functionalized nanoparticles were used as a negative control. Energy dispersive X-ray spectroscopy (EDS) was performed to assay for nitrogen content from the ethylenediamine.

4.3.6 Dendritic Cell Culture and Stimulation

Bone marrow was harvested from C57BL/6 mice purchased from Harlan Laboratories

(Indianapolis, IN). Specific pathogen-free conditions were used to house the mice where all bedding, caging, and feed were sterilized before being used. Approval by the Iowa State

University Institutional Animal Care and Use Committee was received before conducting animal procedures. Bone marrow derived dendritic cells were cultured and differentiated as previously described (22; 32; 33). Bone marrow cells were cultured in RPMI medium containing 1% v/v L- glutamine, 1% v/v penicillin/streptomycin, 0.1% v/v β-mercaptoethanol, 10% v/v heat- inactivated fetal bovine serum and 10 ng/mL granulocyte-monocyte colony stimulating factor until day 10 of culture. After 10 days of culture a small sample of the population was analyzed via flow cytometry and confirmed to be at least 80% CD11c positive. A final cell count was carried out and one million cells were placed into flat bottom 24-well plates. On the 11th day of culture, the DCs were stimulated with 125 μg/mL of QD-loaded or blank (i.e., no QDs) nanoparticles. Each nanoparticle suspension was briefly sonicated at 40 Hz to evenly disperse the aggregated particles. DCs treated with 200 ng/mL LPS were used as a positive control, while non-stimulated DCs served as a negative control. After 48 h of stimulation, cells were harvested

100

for flow cytometric analysis and cytokine secretion. During the internalization experiments,

“false positives” were accounted for by incubating QD-loaded nanoparticles in culture medium for 48 h, followed by centrifugation to pellet the nanoparticles and then exposing the supernatants to cells, as described previously (34).

4.3.7 Flow Cytometric Analysis and Cytokine Secretion Assays

Expression of cell surface markers MHC II, CD206, and CD86 was analyzed using a

Becton-Dickinson FACSAria III flow cytometer (San Jose, CA) and the raw fcs files were analyzed with FlowJo software (TreeStar Inc., Ashland, OR). Cell-free supernatants were collected and analyzed using a BioRad Bio-Plex 200 system (Hercules, CA) for the cytokines IL-

6, IL-1β, TNF-α, and IL-12p40.

4.3.8 Serum Protein Adsorption onto Nanoparticles

Nanoparticles were suspended in a 0.1 M phosphate buffered saline (PBS) solution at pH

7.4 at a concentration of 13.3% w/v. C57BL/6 mouse serum was purchased from Innovative

Research (Novi, MI) and was stored in aliquots at -80 °C until use. A 1:4 volume ratio of nanoparticle suspension to serum was incubated for 30 min at 37 °C in a shaker to achieve adsorption. Non-serum coated particles were incubated in PBS for the same duration under the same conditions to serve as a control. Following incubation, the particles were pelleted from the suspension by centrifugation at 12,000 rcf for 10 min. The supernatant was removed and replaced by an equal volume of PBS. Non-adsorbed proteins were removed from the nanoparticles by pelleting the nanoparticle suspension once more by centrifuging at 12,000 rcf for 10 min. The process of pelleting nanoparticles and resuspending them in PBS was repeated thrice. After the last wash, nanoparticles were dried in a vacuum oven for one hour and stored at

-20 °C until further use.

101

4.3.9 Characterization of Protein Adsorption

Proteins adsorbed to the nanoparticles were quantified and characterized using a bicinchoninic acid (BCA) assay and 2-D gel electrophoresis following elution. 10 mg of dried nanoparticles were suspended in 250 μL of elution buffer for 10 min at 95 °C to elute the adsorbed serum proteins. The elution buffer consisted of a solution of 10% w/v sodium dodecyl sulfate and 2.3% w/v dithiothreitol. The nanoparticles were separated from the eluted proteins in the elution buffer by centrifugation at 12000 rcf for 10 min. This protocol has been shown in previous studies to remove nearly all of the adsorbed proteins from particles (20; 21). The eluted protein solution was passed through a detergent removal column (Thermo Scientific, Rockford,

IL) to remove the sodium dodecyl sulfate that would impede isoelectric focusing during 2-D gel electrophoresis. Portions of the eluted protein solution used in the BCA assay were dialyzed against PBS using a 3 kDa MWCO spin column (Corning, NY) to remove the dithiothreitol that would otherwise interfere with the reagents of the BCA assay. Equal volumes of the eluted protein solution were then used in a BCA protein assay kit to determine the amount of proteins adsorbed to the nanoparticles.

Additionally equal volumes of the eluted protein solution from the 20:80 CPTEG:CPH nanoparticles were loaded onto 7 cm IPG strips, pH 3-10, and underwent active rehydration where the proteins were allowed to diffuse into the IPG strip during a voltage ramping protocol that included 50V for 12 h, 500V for 1 h, 1000V for 1 h, and 8000V for 8 h. This was carried out using an Ettan IPGPhor III system (GE Healthcare, Piscataway, NJ). The set point for current was 50 μA. The eluted protein solution from the 20:80 CPH:SA nanoparticles required passive rehydration for the proteins to separate by isoelectric points. Equal volumes of eluted protein solution from 20:80 CPH:SA nanoparticles were loaded onto 7 cm IPG strips, pH 3-10, and were then coated in mineral oil for 12 h to prevent drying during the passive rehydration process.

102

Subsequently, the rehydrated IPG strips underwent the same voltage ramping protocol as described before.

The second dimension of separation was carried out by loading the IPG strips onto 7 cm

4-20% Tris-HCl polyacrylamide gels for 2.5 h at 110V. Also loaded into the gel were Precision

Plus protein standards from BioRad which included molecular weight standards ranging from

250 kDa to 10 kDa. The gels were then placed in a fixative solution, which contained 40% v/v ethanol and 10% v/v acetic acid, for 3 h at 4 °C. Subsequently the gels were stained with a

BioRad fluorescent flamingo gel stain for 8 h at 4 °C. Nanopure water was used to wash the gels to reduce background fluorescence and placed on a Typhoon 8600 (GE Healthcare) to scan for fluorescence. Progenesis SameSpots software (Nonlinear Dynamics, Durham, NC) was used to identify the molecular weight and isoelectric point of each fluorescent protein spot. The fluorescent intensity of each protein spot was quantified using ImageQuantTL. After accounting for the fluorescent intensity of the entire gel, the individual protein spots were presented as a percentage of the total fluorescent intensity.

4.3.10 Statistical Analysis

Least squared means analysis using a Tukey-Kramer correction to adjust for multiple comparisons was performed to determine statistical significance among the treatments. A p-value less than 0.05 was considered significant.

4.4 Results

4.4.1 Nanoparticle characterization

Particle morphology and size of non-functionalized and functionalized nanoparticles were analyzed using scanning electron microscopy (Table 4.1 and Figure 4.1B). Consistent with

103

previous work, the 20:80 CPH:SA nanoparticles were slightly larger than the 20:80 CPTEG:CPH nanoparticles (35). There were no significant differences in the size of the functionalized particles compared to the corresponding non-functionalized particles (data not shown), which is also consistent with previous work (15).

Nanoparticles were analyzed to determine their ζ potential, carbohydrate content, and elemental composition (Table 4.1). The QELS data showed that the non-functionalized nanoparticles had strongly negative ζ potentials (-23 to -31 mV) due to the deprotonated carboxylic acid end groups of the polyanhydrides. In contrast, the functionalized nanoparticles exhibited a positive charge (8 to 16 mV) due to the protonated amines on the linker. The phenol- sulfuric acid assay was used to confirm the presence of sugar on the nanoparticles, which showed that the functionalized 20:80 CPH:SA nanoparticles had slightly higher sugar densities than their

20:80 CPTEG:CPH counterparts. EDS was used to determine the nanoparticle surface elemental composition. The linker-functionalized particles exhibited greater nitrogen content than the di- mannose particles even though the ζ potentials for both types of particles were relatively similar.

As expected, the EDS data also showed that the oxygen content on the 20:80 CPTEG:CPH nanoparticles was higher than that on the 20:80 CPH:SA nanoparticles.

4.4.2 Serum protein adsorption influenced 20:80 CPH:SA nanoparticle internalization

To determine how nanoparticle chemistry, functionalization, and serum protein adsorption affect interactions with DCs, we first measured the ability of DCs to internalize QD- loaded nanoparticles. At 48 hours post-incubation with nanoparticles, DCs incubated with 20:80

CPH:SA nanoparticles demonstrated a significantly higher percentage of nanoparticle internalization as compared to 20:80 CPTEG:CPH nanoparticles (Figure 4.2). Pre-treating the non-functionalized 20:80 CPH:SA nanoparticles with mouse serum significantly increased the

104

percentage of nanoparticle-positive DCs, however, serum protein adsorption had no effect on the internalization of the 20:80 CPTEG:CPH nanoparticles (closed vs. open bars). Nanoparticle functionalization, either with linker alone or linker plus di-mannose, regardless of nanoparticle chemistry, had no effect on DC internalization. Altogether, these data suggest that both chemistry and serum protein adsorption can affect particle internalization by DCs. Based on these findings, for the remainder of the experiments presented in this manuscript, we utilized nanoparticles pre-incubated with mouse serum in order to better mimic the conditions under which DCs and nanoparticles would interact in vivo.

4.4.3 DC surface marker expression is differentially regulated by chemistry and functionalization

Following encounter with antigen, DCs not only internalize antigen, but undergo a process of maturation in order to acquire the ability to initiate adaptive immune responses. Part of this maturation process is the ability to increase the cell surface expression of proteins that facilitate the interaction with and activation of T cells (6). In order to further characterize how nanoparticle chemistry and functionalization affect the maturation of DCs, we measured the surface expression levels of major histocompatibility complex (MHC) II and CD86 on DCs.

Using flow cytometry, we assessed the expression of these surface markers following a 48 hour incubation with the different nanoparticle formulations. The non-functionalized 20:80

CPTEG:CPH nanoparticles induced significantly higher levels of MHC II and CD86 expression on DCs as compared to DCs incubated with the non-functionalized 20:80 CPH:SA nanoparticles

(Figure 4.3A and 4.3B, non-functionalized nanoparticles). Functionalization of the 20:80

CPTEG:CPH nanoparticles did not affect the level of MHC II surface expression on DCs (Figure

4.3A, black bars) but did significantly increase the surface expression of CD86 on DCs as compared to non-functionalized 20:80 CPTEG:CPH nanoparticles (Figure 4.3B, black bars).

105

DCs incubated with functionalized 20:80 CPH:SA nanoparticles demonstrated significantly increased surface expression of both MHC II and CD86 as compared to incubation with non- functionalized 20:80 CPH:SA nanoparticles (Figure 4.3A and 4.3B, open bars). Altogether, these data demonstrate that polymer hydrophobicity (i.e., the CPH-rich formulation) and the surface modification of less hydrophobic nanoparticles by linker or di-mannose will up-regulate the expression of cell surface markers on DCs.

Stimulation of DCs and macrophages with di-mannose has been previously shown to increase the surface expression of CD206, the mannose receptor (14; 15). To determine if the expression of this receptor was modulated by the functionalization of our nanoparticles, we measured surface CD206 expression on DCs in our experiments. Chemistry did not differentially affect CD206 surface expression (Figure 4.3C, non-functionalized). Regardless of chemistry, linker and di-mannose modified particles induced significantly higher levels of CD206 surface expression with the di-mannose modification resulting in the highest level of CD206 expression

(Figure 4.3C). These findings are consistent with previous studies that demonstrated increased expression of this receptor in the presence of mannosylated ligands (14; 15).

4.4.4 DC cytokine production is differentially regulated by nanoparticle chemistry and functionalization

The development and skewing of adaptive immune responses are also dependent on the production of cytokines by DC and other antigen presenting cells (36). In order to determine how the nanoparticle formulations impact cytokine production, cytokine secretion from culture supernatants of DCs incubated with serum-coated nanoparticles was analyzed. DCs incubated with non-functionalized 20:80 CPH:SA nanoparticles secreted significantly higher amounts of

IL-6, IL-1, and TNF- as compared to DCs incubated with non-functionalized 20:80

106

CPTEG:CPH nanoparticles (Figure 4.4). However, functionalization had differing effects on cytokine production depending on chemistry. Linker- and di-mannose-functionalized 20:80

CPTEG:CPH nanoparticles resulted in significantly increased secretion of IL-6, IL-1, and TNF-

 by DCs as compared to non-functionalized 20:80 CPTEG:CPH nanoparticles (Figure 4.4A,

4.4B and 4.4C, black bars). In contrast, functionalization of 20:80 CPH:SA nanoparticles resulted in a decrease in cytokine production by DCs as compared to 20:80 CPH:SA non- functionalized nanoparticles (Figure 4.4A, 4.4B and 4.4C, open bars). Altogether, these data demonstrate that while chemistry does play a role in the induction of cytokine production by

DCs, carbohydrate surface modification can modulate such responses in a chemistry-dependent manner.

4.4.5 Adsorption of immune activating serum proteins onto polyanhydride nanoparticles

The data presented so far suggest that nanoparticle chemistry and surface modification differentially regulate particle internalization, surface expression of cell surface markers, and cytokine production by DCs. As we have previously demonstrated, nanoparticle hydrophobicity affects adsorption of serum proteins and results in differential interactions with DCs (20; 37; 38).

Together with hydrophobicity, surface functionalization can also affect the interaction between the particles and serum proteins. To better understand the differential effects of hydrophobicity and surface functionalization on interactions with serum proteins, we assessed the amount and type of proteins that are adsorbed by the different nanoparticle formulations following incubation with serum. The 20:80 CPTEG:CPH nanoparticles adsorbed a significantly greater amount of serum protein, which was at least five times higher compared to that adsorbed by the 20:80

CPH:SA nanoparticles (Figure 4.5). Linker and di-mannose modification of the 20:80

CPTEG:CPH nanoparticles resulted in significantly higher amounts of adsorbed serum proteins,

107

as compared to non-functionalized 20:80 CPTEG:CPH nanoparticles (Figure 4.5, last 3 bars). In contrast, functionalizing the 20:80 CPH:SA nanoparticles with linker or di-mannose had only a modest impact on protein adsorption compared to the 20:80 CPTEG:CPH formulation (Figure

4.5, first 3 bars).

The eluted serum proteins were further characterized using 2-D gel electrophoresis

(Figure 4.6). Based on this analysis, albumin and IgG were the most abundant proteins adsorbed onto both nanoparticle formulations (Table 4.2). In addition, there were a large number of immune activating proteins adsorbed onto the nanoparticles. The predominant candidates included IgG and complement component C3, which was represented by the two proteins C3a and C3b (39; 40).

Overall, the 20:80 CPTEG:CPH nanoparticle formulation demonstrated a greater amount of serum protein adsorption. On the 2-D gels, eluted serum proteins from 20:80 CPTEG:CPH nanoparticles (Figure 4.6) showed a greater diversity of proteins adsorbed as compared to the

20:80 CPH:SA nanoparticles (Figure 4.6 and Table 4.2A and 4.2B). Independent of nanoparticle chemistry, functionalizing with di-mannose resulted in adsorption of mannose binding protein C, which was not detected in the serum eluted from the non-functionalized nanoparticles.

4.5 Discussion

Initiation of adaptive immune responses requires antigen capture and processing by DCs and the subsequent maturation of these cells as they traffic from the site of antigen encounter to the draining lymph node (41). During this process, DCs up-regulate surface expression of MHC

II, necessary for antigen presentation to T cells (i.e., signal 1), and surface expression of co- stimulatory molecules such as CD86 (i.e., signal 2). In addition, DCs secrete cytokines (i.e., signal 3), which are necessary to enhance and skew the adaptive immune response (36). In this

108

work, we demonstrate that nanoparticle chemistry and surface functionalization differentially regulated DC maturation and polyfunctionality (i.e., cell surface marker expression and cytokine secretion). Understanding how these factors affect DC maturation can lead to rational nanoparticle-based vaccine design, which targets specific aspects of DC maturation as well as the nature and magnitude of the subsequent immune response. Moreover, this work demonstrates how polyanhydride nanoparticles function as more than just vaccine-delivery vehicles.

Nanoparticle chemistry and surface functionalization, prior to antigen delivery, can serve as modulators of antigen presenting cell responses, thus tailoring the ensuing adaptive immune response.

As indicated in Table 4.1, the functionalized 20:80 CPH:SA nanoparticles had slightly higher sugar densities than the functionalized 20:80 CPTEG:CPH nanoparticles. This could be attributed to the higher glass transition temperature of the 20:80 CPH:SA nanoparticles, which makes these particles less likely to aggregate and block functionalization (23). The EDS data in

Table 4.1 indicates that the glycolic acid-functionalized particles have higher nitrogen content than the di-mannose particles even though the ζ potentials for both types of particles were relatively similar. This could be attributed to the larger di-mannose being richer in carbon and oxygen, which reduces the nitrogen content. Likewise, the higher oxygen content on the surface of the 20:80 CPTEG:CPH nanoparticles compared to that on the 20:80 CPH:SA nanoparticles can be attributed to the presence of ethylene glycol moieties in the CPTEG backbone.

Our initial studies demonstrated that nanoparticle chemistry alone plays a major role in determining the level of nanoparticle internalization by DCs in vitro. The less hydrophobic 20:80

CPH:SA formulation was internalized at significantly higher levels as compared to the more hydrophobic 20:80 CPTEG:CPH nanoparticles (Figure 4.2). Prior incubation with mouse serum

109

enhanced the internalization of 20:80 CPH:SA nanoparticles, but had no effect on the internalization of 20:80 CPTEG:CPH nanoparticles. This result demonstrates that nanoparticle chemistry is an important aspect of vaccine design, as serum protein adsorption alone may or may not change the biological effect(s) in vivo. Functionalization did not significantly affect nanoparticle internalization by DCs regardless of chemistry or the presence of serum. These data suggest that surface modification is not sufficient to alter the initial interactions between DCs and nanoparticles, and thus, does not have a major effect on nanoparticle internalization.

To further characterize how nanoparticle chemistry and functionalization can modulate

DC maturation and function, we assessed DC surface marker expression and cytokine production profiles. Despite the lower DC internalization levels, the non-functionalized 20:80 CPTEG:CPH nanoparticles induced significantly higher surface expression of MHC II and CD86 than did the non-functionalized 20:80 CPH:SA nanoparticles (Figure 4.3). This bystander effect has been observed previously, specifically for MHC II and CD86, where DC internalization of polyanhydride nanoparticles alone did not directly correlate with enhanced activation (34). These findings are also consistent with previous reports that suggest the importance of hydrophobicity as a danger signal that triggers immune activation (34; 42). The cell surface marker expression data presented in this work reinforces this concept as MHC II and CD86 expression on DCs were up-regulated most strongly when stimulated by the more hydrophobic non-functionalized 20:80

CPTEG:CPH nanoparticles.

While the hydrophobic 20:80 CPTEG:CPH nanoparticles induced higher expression of

MHC II and CD86, nanoparticle functionalization had no effect on MHC II surface expression levels regardless of chemistry (Figure 4.3A). This data would suggest that carbohydrate modification of nanoparticles and engagement of CLRs does not affect MCH II expression.

110

However, functionalization did increase surface expression of CD86 (Figure 4.3B), as we observed a step-wise increase in CD86 expression by stimulated DCs: non-functionalized < linker- < di-mannose-functionalized nanoparticles. As surface modification increases the complexity (i.e., more receptor-ligand binding) of the particle-DC interaction, it may enhance or sustain intracellular signals necessary to induce more robust MHC II and CD86 expression.

These findings suggest that the increased expression of MHC II (i.e., signal 1, antigen presentation) was more dependent on nanoparticle chemistry than surface functionalization of the nanoparticles. In contrast, expression of co-stimulatory molecules such as CD86 (i.e., signal

2), while also dependent on chemistry, was further modulated by surface functionalization.

Cytokine secretion by DCs is a key factor in promoting and skewing subsequent adaptive immune responses, also known as signal 3. When assessing cytokine expression profiles of DCs stimulated with the different nanoparticle formulations, we observed a differential pattern of cytokine secretion. In the presence of serum proteins, the less hydrophobic non-functionalized

20:80 CPH:SA nanoparticles induced the secretion of significantly higher amounts of IL-6, IL-

1, and TNF- as compared to the non-functionalized 20:80 CPTEG:CPH nanoparticles (Figure

4.4). This is unlike the pattern we observed for surface marker expression (Figure 4.3), where the more hydrophobic 20:80 CPTEG:CPH nanoparticles resulted in higher surface expression of these molecules. This would suggest that in terms of cytokine production by DCs the bystander effect (i.e., particle negative DCs) induced by the non-functionalized 20:80 CPTEG:CPH nanoparticles plays a lesser role as compared to surface receptor expression.

Interestingly, nanoparticle functionalization resulted in opposite effects on cytokine expression depending on nanoparticle chemistry (Figure 4.4). For the 20:80 CPH:SA nanoparticles, functionalization resulted in a negative regulation of cytokine production for IL-6,

111

IL-1, and TNF-. In contrast, functionalization of the 20:80 CPTEG:CPH nanoparticles led to the up-regulation of cytokine release by DCs, resulting in significantly higher amounts of IL-6,

IL-1, and TNF- as compared to non-functionalized 20:80 CPTEG:CPH nanoparticles. This suggests that the functionalized 20:80 CPTEG:CPH nanoparticles would induce a more robust signal 3, likely through a bystander effect, which is consistent with previous studies that showed that the 20:80 CPTEG:CPH formulation induced more robust T cell responses (43). Altogether, these data demonstrate that nanoparticle chemistry itself can differentially regulate DC cytokine production. More importantly, our data also shows that surface modification of polyanhydride nanoparticles can affect cytokine production in a chemistry-dependent manner. For completeness, the cell surface marker expression and cytokine secretion data for the non-serum- adsorbed particles are shown in the Supporting Information section.

Dendritic cell maturation is triggered by signals received during its interactions with the antigen or pathogen and the environmental milieu in which encounter occurs. Some of these interactions are mediated by plasma membrane receptors including PRRs, CLRs, Fc gamma receptors, and complement (C) receptors (44). Some of these receptors recognize foreign molecules while others recognize host proteins, namely serum proteins. Serum proteins adsorbed onto our nanoparticles will likely modulate the way the nanoparticles interact with DCs.

Moreover, we hypothesize that chemistry and surface modification of the nanoparticles would lead to differential interactions with serum proteins and may contribute to the differences in internalization, DC surface receptor expression, and cytokine production. Following analysis of adsorbed serum protein onto our different nanoparticle formulations, we determined (as expected) that the more hydrophobic 20:80 CPTEG:CPH nanoparticles adsorbed two-to-four fold greater amounts of protein regardless of functionalization as compared to the 20:80 CPH:SA

112

nanoparticles (Figure 4.5). In addition, linker- and di-mannose functionalization of the nanoparticles increased the levels of adsorbed protein as compared to the non-functionalized nanoparticles. This increase in the amount of serum protein adsorption upon functionalization suggests that in addition to hydrophobicity, electrostatic interactions may also affect protein adsorption.

The 2-D gel electrophoresis data (Figure 4.6) further confirmed the polymer chemistry- dependent effects on serum adsorption. Not only did the more hydrophobic 20:80 CPTEG:CPH particles adsorb more total serum proteins, they also adsorbed a more diversified profile of proteins. The characterization of the eluted proteins shown in Table 4.2 indicated that serum- abundant protein such as albumin and IgG were adsorbed onto the nanoparticles, regardless of chemistry.

Based on the 2D gel electrophoresis analysis, neither linker nor sugar functionalization appeared to have a major effect on the profile of adsorbed serum proteins. However, sugar functionalization enhanced the adsorption of mannose-binding protein, especially on the 20:80

CPTEG:CPH nanoparticles (Table 4.2). This is an interesting observation, suggesting that the presence of the ligand alone is not enough to bind mannose-binding protein, and that other aspects including hydrophobicity, presence of oxygen moieties, and electrostatic interactions may also lead to enhanced binding profiles. The presence of the mannose binding protein C in the serum eluted from the functionalized nanoparticles is significant because the mannose binding protein C is a complement-activating protein that may aid in the enhanced activation of

DCs (45).

The complement protein, C3a, adsorbed onto the 20:80 CPTEG:CPH but not the 20:80

CPH:SA nanoparticle formulation. C3a is produced from the enzymatic cleavage of C3 and is

113

characterized as an anaphylatoxin, a potent inducer of inflammation, which acts via binding onto

C3a receptors (40; 46). Binding of C3a onto 20:80 CPTEG:CPH nanoparticles may lead to enhanced inflammatory signals by cells interacting with the particles. Surface modification either by linker or di-mannose did not markedly alter the level of C3a adsorption, suggesting functionalization does not greatly affect the way C3a interacts with the CPTEG:CPH nanoparticles. The absence of C3a on 20:80 CPH:SA nanoparticles suggests this formulation may not provide efficient binding conditions for adsorption and points to the use of this formulation in situations where C3a-mediated responses are not desired. We also observed binding of C3b, generated from the enzymatic cleavage of C3 (40). C3b is a potent opsonin, whose action is mediated by C3b receptors found on DC plasma membranes (47). We did not observe any marked differences in C3b adsorption between chemistries or following functionalization. In addition to hydrophobicity, C3b, given its function, could partly mediate the level of particle internalization. However, by itself, C3b does not appear to be the major driving force behind the previously observed dichotomy between CPTEG:CPH (low) and CPH:SA

(high) internalization by DCs.

While both nanoparticle formulations adsorbed serum proteins, the differences in total protein amount cannot be directly correlated to DC maturation. This is not surprising given the multi-factorial nature of DC-nanoparticle interactions. Moreover, while we did observe differences in the bound proteins, namely mannose-binding lectin and C3a, again, these differences do not provide a direct mechanism by which the differential effects on DC maturation occur with our nanoparticle formulations. However, these data demonstrate that nanoparticle-serum protein interactions must be taken into account during the rational design of nanoparticle-based vaccines.

114

4.6 Conclusions

Our results show that polyanhydride nanoparticle chemistry, surface functionalization of the nanoparticles by carbohydrates, and serum protein adsorption influenced nanoparticle interactions with DCs. In this study, we characterized the DC response to nanoparticle stimulation in terms of internalization ability, MHC II expression (signal 1), up-regulation of co- stimulatory molecule (CD86; signal 2) and cytokine production (signal 3). The events described here are critical for antigen recognition and for the subsequent induction of an adaptive immune response. In this work, we demonstrated that the more hydrophobic CPTEG:CPH nanoparticle formulation was not internalized with high efficiency and did not induce production of IL-6, IL-

1, or TNF- by DCs. However, this formulation enhanced the surface expression of MHC II and CD86. In contrast, the less hydrophobic CPH:SA nanoparticle formulation was efficiently internalized by DCs, induced higher amounts of IL-6, IL-1, and TNF-secretion, but did not trigger surface expression of MCH II or CD86. While functionalization of either chemistry by linker or di-mannose did not change internalization efficiency or surface maker expression levels in the presence of serum, it resulted in differing effects on cytokine production, in a chemistry- dependent manner. Functionalization of the CPTEG:CPH nanoparticles enhanced cytokine secretion, while functionalization of the CPH:SA nanoparticles led to a decrease in cytokine production. These observations enhance our understanding of how non-functionalized and carbohydrate-functionalized particles could interact with immune cells in vivo and of the need to consider the impact of nanoparticles on host cells in the context of adsorbed serum proteins.

Such an understanding will aid in the rational design of nanoparticle-based vaccines.

115

4.7 Acknowledgments

The authors would like to acknowledge financial support from the U.S. Army Medical

Research and Materiel Command (Grant No. W81XWH-10-1-0806). The authors would also like to thank Shawn Rigby of the Iowa State Flow Cytometry facility, Joel Nott of the Iowa State

University Protein Facility for his assistance with the 2-D gel electrophoresis experiments and documentation, and Dr. Yaxuan Sun of Iowa State’s Department of Statistics for her help with statistical analysis of the data.

4.8 References 1. Petersen LK, Sackett CK, Narasimhan B. 2010. High-throughput analysis of protein stability in polyanhydride nanoparticles. Acta biomaterialia 6:3873-81

2. Petersen LK, Sackett CK, Narasimhan B. 2010. Novel, high throughput method to study in vitro protein release from polymer nanospheres. Journal of combinatorial chemistry 12:51-6

3. Kipper MJ, Wilson JH, Wannemuehler MJ, Narasimhan B. 2006. Single dose vaccine based on biodegradable polyanhydride microspheres can modulate immune response mechanism. Journal of biomedical materials research. Part A 76:798-810

4. Huntimer L, Wilson Welder JH, Ross K, Carrillo-Conde B, Pruisner L, et al. 2013. Single immunization with a suboptimal antigen dose encapsulated into polyanhydride microparticles promotes high titer and avid antibody responses. Journal of biomedical materials research. Part B, Applied biomaterials 101:91-8

5. Ulery BD, Phanse Y, Sinha A, Wannemuehler MJ, Narasimhan B, Bellaire BH. 2009. Polymer chemistry influences monocytic uptake of polyanhydride nanospheres. Pharm Res 26:683-90

6. van Vliet SJ, Garcia-Vallejo JJ, van Kooyk Y. 2008. Dendritic cells and C-type lectin receptors: coupling innate to adaptive immune responses. Immunology and cell biology 86:580-7

7. Mallapragada SK, Narasimhan B. 2008. Immunomodulatory biomaterials. International journal of pharmaceutics 364:265-71

116

8. Figdor CG, van Kooyk Y, Adema GJ. 2002. C-type lectin receptors on dendritic cells and Langerhans cells. Nature reviews. Immunology 2:77-84

9. van Kooyk Y, Rabinovich GA. 2008. Protein-glycan interactions in the control of innate and adaptive immune responses. Nature immunology 9:593-601

10. Rudd PM, Elliott T, Cresswell P, Wilson IA, Dwek RA. 2001. Glycosylation and the immune system. Science 291:2370-6

11. Adams EW, Ratner DM, Seeberger PH, Hacohen N. 2008. Carbohydrate-mediated targeting of antigen to dendritic cells leads to enhanced presentation of antigen to T cells. Chembiochem : a European journal of chemical biology 9:294-303

12. Garcia-Vallejo JJ, Ambrosini M, Overbeek A, van Riel WE, Bloem K, et al. 2013. Multivalent glycopeptide dendrimers for the targeted delivery of antigens to dendritic cells. Molecular immunology 53:387-97

13. Phanse Y, Carrillo-Conde BR, Ramer-Tait AE, Roychoudhury R, Pohl NL, et al. 2013. Functionalization of polyanhydride microparticles with di-mannose influences uptake by and intracellular fate within dendritic cells. Acta biomaterialia

14. Carrillo-Conde B, Song EH, Chavez-Santoscoy A, Phanse Y, Ramer-Tait AE, et al. 2011. Mannose-functionalized "pathogen-like" polyanhydride nanoparticles target C-type lectin receptors on dendritic cells. Molecular pharmaceutics 8:1877-86

15. Chavez-Santoscoy AV, Roychoudhury R, Pohl NL, Wannemuehler MJ, Narasimhan B, Ramer-Tait AE. 2012. Tailoring the immune response by targeting C-type lectin receptors on alveolar macrophages using "pathogen-like" amphiphilic polyanhydride nanoparticles. Biomaterials 33:4762-72

16. Wijagkanalan W, Kawakami S, Takenaga M, Igarashi R, Yamashita F, Hashida M. 2008. Efficient targeting to alveolar macrophages by intratracheal administration of mannosylated liposomes in rats. Journal of controlled release : official journal of the Controlled Release Society 125:121-30

17. Song EH, Osanya AO, Petersen CA, Pohl NL. 2010. Synthesis of multivalent tuberculosis and Leishmania-associated capping carbohydrates reveals structure-dependent responses allowing immune evasion. Journal of the American Chemical Society 132:11428-30

117

18. Engering AJ, Cella M, Fluitsma D, Brockhaus M, Hoefsmit EC, et al. 1997. The mannose receptor functions as a high capacity and broad specificity antigen receptor in human dendritic cells. European journal of immunology 27:2417-25

19. Tacken PJ, de Vries IJ, Torensma R, Figdor CG. 2007. Dendritic-cell immunotherapy: from ex vivo loading to in vivo targeting. Nature reviews. Immunology 7:790-802

20. Carrillo-Conde B, Garza A, Anderegg J, Narasimhan B. 2010. Protein adsorption on biodegradable polyanhydride microparticles. Journal of biomedical materials research. Part A 95:40-8

21. Carrillo-Conde BR, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2012. Chemistry-dependent adsorption of serum proteins onto polyanhydride microparticles differentially influences dendritic cell uptake and activation. Acta biomaterialia 8:3618-28

22. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

23. Torres MP, Vogel BM, Narasimhan B, Mallapragada SK. 2006. Synthesis and characterization of novel polyanhydrides with tailored erosion mechanisms. Journal of biomedical materials research. Part A 76:102-10

24. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

25. Huntimer L, Ramer-Tait AE, Petersen LK, Ross KA, Walz KA, et al. 2013. Evaluation of biocompatibility and administration site reactogenicity of polyanhydride-particle-based platform for vaccine delivery. Advanced healthcare materials 2:369-78

26. Vogel BM, Mallapragada SK. 2005. Synthesis of novel biodegradable polyanhydrides containing aromatic and glycol functionality for tailoring of hydrophilicity in controlled drug delivery devices. Biomaterials 26:721-8

27. Determan AS, Wilson JH, Kipper MJ, Wannemuehler MJ, Narasimhan B. 2006. Protein stability in the presence of polymer degradation products: consequences for controlled release formulations. Biomaterials 27:3312-20

118

28. Shen E, Pizsczek R, Dziadul B, Narasimhan B. 2001. Microphase separation in bioerodible copolymers for drug delivery. Biomaterials 22:201-10

29. Conix A. 1966. Poly[1,3-bis(p-carboxyphenoxy)-propane anhydride]. Macromolecular Synthesis 2:95-8

30. Jaipuri FA, Pohl NL. 2008. Toward solution-phase automated iterative synthesis: fluorous-tag assisted solution-phase synthesis of linear and branched mannose oligomers. Organic & biomolecular chemistry 6:2686-91

31. Masuko T, Minami A, Iwasaki N, Majima T, Nishimura S, Lee YC. 2005. Carbohydrate analysis by a phenol-sulfuric acid method in microplate format. Analytical biochemistry 339:69-72

32. Lutz MB, Kukutsch N, Ogilvie AL, Rossner S, Koch F, et al. 1999. An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. Journal of immunological methods 223:77-92

33. Boggiatto PM, Jie F, Ghosh M, Gibson-Corley KN, Ramer-Tait AE, et al. 2009. Altered dendritic cell phenotype in response to Leishmania amazonensis amastigote infection is mediated by MAP kinase, ERK. The American journal of pathology 174:1818-26

34. Petersen LK, Ramer-Tait AE, Broderick SR, Kong CS, Ulery BD, et al. 2011. Activation of innate immune responses in a pathogen-mimicking manner by amphiphilic polyanhydride nanoparticle adjuvants. Biomaterials 32:6815-22

35. Petersen LK, Phanse Y, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2012. Amphiphilic polyanhydride nanoparticles stabilize Bacillus anthracis protective antigen. Molecular pharmaceutics 9:874-82

36. Blanco P, Palucka AK, Pascual V, Banchereau J. 2008. Dendritic cells and cytokines in human inflammatory and autoimmune diseases. Cytokine & growth factor reviews 19:41-52

37. Petersen LK, Xue L, Wannemuehler MJ, Rajan K, Narasimhan B. 2009. The simultaneous effect of polymer chemistry and device geometry on the in vitro activation of murine dendritic cells. Biomaterials 30:5131-42

119

38. Petersen LK, Huntimer L, Walz K, Ramer-Tait A, Wannemuehler MJ, Narasimhan B. 2013. Combinatorial evaluation of in vivo distribution of polyanhydride particle-based platforms for vaccine delivery. International journal of nanomedicine 8:2213-25

39. Duan X, Yarmush DM, Berthiaume F, Jayaraman A, Yarmush ML. 2004. A mouse serum two-dimensional gel map: application to profiling burn injury and infection. Electrophoresis 25:3055-65

40. Sahu A, Lambris JD. 2001. Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunological reviews 180:35-48

41. Randolph GJ, Angeli V, Swartz MA. 2005. Dendritic-cell trafficking to lymph nodes through lymphatic vessels. Nature reviews. Immunology 5:617-28

42. Seong SY, Matzinger P. 2004. Hydrophobicity: an ancient damage-associated molecular pattern that initiates innate immune responses. Nature reviews. Immunology 4:469-78

43. Huntimer LM, Ross KA, Darling RJ, Winterwood NE, Boggiatto P, et al. 2014. Polyanhydride nanovaccine platform enhances antigen-specifi c cytotoxic T cell responses. Technology 2:171-5

44. Palm NW, Medzhitov R. 2009. Pattern recognition receptors and control of adaptive immunity. Immunological reviews 227:221-33

45. Turner MW. 1996. Mannose-binding lectin: the pluripotent molecule of the innate immune system. Immunology today 17:532-40

46. Monsinjon T, Gasque P, Ischenko A, Fontaine M. 2001. C3A binds to the seven transmembrane anaphylatoxin receptor expressed by epithelial cells and triggers the production of IL-8. FEBS letters 487:339-46

47. Sarma JV, Ward PA. 2011. The complement system. Cell and tissue research 343:227-35

120

4.9 Figures

Figure 4.1 Polyanhydride chemistry, nanoparticle preparation, and surface functionalization. A) Structure of monomer diacids: sebacic acid (top left), 1,6-bis(p-carboxyphenoxy)hexane (top right) and 1,8-bis(p- carboxyphenoxy)-3,6-dioxaoctane (bottom). B) Copolymers are used to synthesize nanoparticles using anti-solvent nanoprecipitation. Scanning electron photomicrographs of 20:80 CPTEG:CPH (left) and

20:80 CPH:SA (right) nanoparticles are shown. C) The nanoparticle surface was functionalized with di- mannose (left) or a glycolic acid cap (right) using an ethylenediamine linker.

121

Figure 4.2 Serum coated nanoparticle internalization by DCs is a strong function of polymer chemistry. Serum coated or non-serum coated QD-loaded nanoparticles were incubated with

DCs for 48 hours before harvesting the DCs and analyzing them by flow cytometry. Data are expressed as the mean ± SEM of two independent experiments performed in triplicate. * indicates statistical significance (p < 0.05) between 20:80 CPH:SA and 20:80 CPTEG:CPH nanoparticles for the same functionalization and serum treatment; # indicates statistical significance (p < 0.05) between serum coated and non-serum coated particles of the same chemistry and functionalization. NF = non-functionalized in Figure.

122

Figure 4.3 Surface functionalization and nanoparticle chemistry influence the expression of (A)

MHC II, (B) CD86, and (C) CD206 when DCs were incubated for 48 hours with serum coated nanoparticles. Data is expressed as the means ± SEM of two independent experiments performed in triplicate. ^ indicates statistical significance (p < 0.05) from non-stimulated cells; + indicates statistical significance (p < 0.05) between functionalized and non-functionalized particles of the same chemistry; * indicates statistical significance (p < 0.05) between 20:80 CPH:SA and 20:80

CPTEG:CPH nanoparticles for the same functionalization and serum treatment. NF = non- functionalized in Figure.

123

Figure 4.4 Pro-inflammatory cytokine secretion from DCs incubated with serum-coated nanoparticles for 48 hours is dependent upon nanoparticle chemistry, and surface functionalization. The cellular supernatants of DCs were assayed for (A) IL-6, (B) IL-1β, and (C)

TNF-α. Data are expressed as the mean ± SEM of two independent experiments performed in triplicate. Stimulation of DCs with LPS served as a positive control and the recovered supernatants had mean cytokine concentrations of 58,362 pg/mL for IL-6, 36 pg/mL for IL-1β, and 23,419 pg/mL for TNF-α. ^ indicates statistical significance (p < 0.05) from non-stimulated

124

cells; + indicates statistical significance (p < 0.05) between functionalized and non- functionalized particles of the same chemistry; * indicates statistical significance (p < 0.05) between 20:80 CPH:SA and 20:80 CPTEG:CPH nanoparticles for the same functionalization and

serum treatment. NF = non-functionalized in Figure.

)

i t

r 5 0

a +

g *

4 0 m

/ + g

 * (

3 0

n *

i t

p 2 0

o s

d 1 0 +

e g

a 0

r e A A A H H H v S : :S :S P P P

A C H H H : :C :C P P P G G G C C C E E E 0 0 0 T T T 8 : :8 :8 P P P 0 0 0 C C C 2 2 2 0 0 0 r 8 F e : :8 :8 N e s 0 k o 0 0 n 2 2 2 i n r L n F e a N e s k o -m in n i L n D a m i- D

Figure 4.5 Adsorption of serum proteins is dependent on nanoparticle chemistry and surface functionalization. Data are presented as the mean ± SEM of three independent experiments performed in triplicate. + indicates statistical significance (p < 0.05) between the functionalized and non-functionalized particles of the same chemistry; * indicates statistical significance (p <

0.05) between 20:80 CPH:SA and 20:80 CPTEG:CPH nanoparticles for the same functionalization and serum treatment. The p-value for the comparison between the serum protein adsorption on di-mannose-functionalized 20:80 CPH:SA nanoparticles and the corresponding non-functionalized particles was 0.0843. NF = non-functionalized in Figure.

125

Figure 4.6 Serum protein adsorption patterns are affected by nanoparticle chemistry and surface functionalization. Shown are representative 2-D gels of adsorbed serum proteins eluted from (A) non-functionalized 20:80 CPTEG:CPH, (B) linker-functionalized 20:80 CPTEG:CPH, (C) di- mannose-functionalized 20:80 CPTEG:CPH, (D) non-functionalized 20:80 CPH:SA, (D) linker- functionalized 20:80 CPH:SA, and (E) di-mannose-functionalized 20:80 CPH:SA nanoparticles.

NF = non-functionalized in Figure.

126

4.10 Tables Table 4.1 Nanoparticle size and characterization. Average Zeta Polydispersity Sugar Densityc EDS Percentagesd Nanoparticle Type Diametera Potentialb Index (µg/mg) (nm) (mV) Average % C Average % N Average % O 20:80 CPTEG:CPH NF 149 ± 33 0.108 -23.7 ± 0.9 83.85 ± 1.41 0.16 ± 0 .11 16.1 ± 1.52 Linker 159 ± 31 0.104 11.9 ± 2.1 79.27 ± 1.35 2.07 ± 0.84 18.09 ± 0.95 Di-mannose 176 ± 31 0.105 8.4 ± 1.7 6.6 ± 1.71 80.14 ± 1.18 0.85 ± 0.51 19.01 ± 1.19 20:80 CPH:SA NF 227 ± 60 0.157 -31 ± 0.5 91.82 ± 1.74 0.24 ± 0.18 8.18 ± 1.75 Linker 242 ± 72 0.163 15.8 ± 1.6 85.13 ± 2.38 4.15 ± 1.24 10.09 ± 1.55 Di-mannose 251 ± 49 0.149 14.1 ± 1.9 12.3 ± 2.01 88.98 ± 1.20 2.2 ± 0.91 8.72 ± 0.85 aNanoparticle sizes were determined using ImageJ v1.36b analysis of scanning electron photomicrographs using at least 200 particles from each image. bZeta potential measurements are the mean ± standard deviation from three independent experiments. cMean ± standard deviation of the sugar density was determined using a phenol-sulfuric acid assay measured in triplicate. dEDS was used to identify an increase in nitrogen content to confirm surface attachment of the sugar and ethylenediamine. This is shown as mean ± standard deviation from five independent experiments.

Table 4.2A Percent fluorescent intensity of each protein spot for 20:80 CPTEG:CPH nanoparticlesa

Fluorescent Intensities Protein Non-functionalized Linker Di-mannose Albumin 29.91% 26.89% 19.06% IgG Heavy Chain 17.26% 14.45% 10.36% IgG Light Chain 9.73% 8.65% 17.39% C3a 5.74% 10.14% 5.97% C3b 6.53% 7.57% 4.07% IgM 9.85% 10.72% 1.78% Apolipoprotein A-I 1.34% 2.97% 5.96% Apolipoprotein E 3.54% 6.25% 5.32% Apolipoprotein H 0.00% 4.45% 4.32% Transferrin 3.87% 0.00% 0.00% Mannose Binding Protein C 0.00% 0.78% 20.22% Apolipoprotein C-III 0.00% 0.00% 0.00% Unidentified Protein Spots 12.23% 7.14% 5.54%

127

Table 4.2B Percent fluorescent intensity of each protein spot for 20:80 CPH:SA nanoparticlesa

Fluorescent Intensities Protein Non-functionalized Linker Di-mannose Albumin 20.04% 31.40% 21.74% IgG Heavy Chain 27.57% 27.89% 13.46% IgG Light Chain 20.28% 13.75% 29.41% C3a 0.00% 0.00% 0.00% C3b 7.98% 10.94% 7.73% IgM 9.49% 2.59% 0.00% Apolipoprotein A-I 0.00% 0.00% 0.00% Apolipoprotein E 0.00% 0.00% 6.05% Apolipoprotein H 4.12% 0.00% 0.00% Transferrin 0.00% 2.00% 0.00% Mannose Binding Protein C 0.00% 4.71% 7.26% Apolipoportein C-III 0.00% 1.09% 3.84% Unidentified Protein Spots 10.53% 5.64% 10.51% aEach protein spot was quantified using ImageQuantTL and normalized to the total fluorescence intensity of the gel. Data is the average of two independent experiments.

128

4.11 Supporting Information

Supporting Figure 4.1 Surface functionalization and nanoparticle chemistry influenced the expression of (A) MHC II, (B) CD86, and (C) CD206 when dendritic cells (DCs) were incubated for 48 hours with non-serum coated nanoparticles. The addition of 20:80 CPH:SA nanoparticles to the culture wells did not statistically affect the surface expression of MHC II on DCs, while linker and di-mannose modification enhanced the expression of CD86 and CD206 on DCs.

Regardless of functionalization, the presence of the 20:80 CPTEG:CPH nanoparticles significantly increased the cell surface expression of MHC II, CD86, and CD206 on DCs. Data is expressed as the mean ± SEM of two independent experiments. ^ indicates statistical significance

129

(p < 0.05) from non-stimulated (NS) cells; + indicates statistical significance (p < 0.05) between functionalized and NF particles of the same chemistry; * indicates statistical significance (p <

0.05) between 20:80 CPH:SA and 20:80 CPTEG:CPH nanoparticles for the same functionalization. NF = non-functionalized in Figure.

Supporting Figure 4.2 For 20:80 CHP:SA and 20:80 CPTEG:CPH nanoparticles, surface functionalization with di-mannose enhanced dendritic cell (DC) secretion of A) IL-6, B) IL-1β, and C) TNF-α. Polymer chemistry had little influence on the cytokine secretion of DCs cultured with non-serum coated linker functionalized particles. Functionalization with di-mannose generally increased cytokine secretion from DCs incubated with non-serum coated particles of

130

both polymer chemistries. Data is expressed as the mean ± SEM of two independent experiments performed in triplicate. ^ indicates statistical significance (p < 0.05) from non-stimulated cells; + indicates statistical significance (p < 0.05) between functionalized and NF particles of the same chemistry; * indicates statistical significance (p < 0.05) between 20:80 CPH:SA and 20:80

CPTEG:CPH nanoparticles for the same functionalization and serum treatment. NF = non- functionalized in Figure.

131

CHAPTER 5: SAFETY AND BIOCOMPATIBILITY OF

CARBOHYDRATE-FUNCTIONALIZED POLYANHYDRIDE

NANOPARTICLES

Julia E. Vela Ramirez1, Jonathan T. Goodman1, Paola M. Boggiatto2, Rajarshi Roychoudhury3,

Nicola L.B. Pohl3, Jesse M. Hostetter4, Michael J. Wannemuehler2, and Balaji Narasimhan1

Reprinted from The AAPS Journal 17(1):256–267, 2015

5.1 Abstract

Carbohydrate functionalization of nanoparticles allows for targeting of C-type lectin receptors. This family of pattern recognition receptors expressed on innate immune cells, such as macrophages and dendritic cells, can be used to modulate immune responses. In this work, the in vivo safety profile of carbohydrate-functionalized polyanhydride nanoparticles was analyzed following parenteral and intranasal administration in mice. Polyanhydride nanoparticles based on

1,6-bis-(p-carboxyphenoxy)hexane and 1,8-bis-(p-carboxyphenoxy)-3,6-dioxaoctane were used.

Nanoparticle functionalization with di-mannose (specifically carboxymethyl-α-D- mannopyranosyl-(1,2)-D-mannopyranoside), galactose (specifically carboxymethyl- - galactoside), or glycolic acid induced no adverse effects after administration based on histopathological evaluation of liver, kidneys, and lungs. Regardless of the polymer formulation, there was no evidence of hepatic or renal damage or dysfunction observed in serum or urine samples. The histological profile of cellular infiltration and the cellular distribution and kinetics in the lungs of mice administered nanoparticle treatments followed similar behavior as that observed in the lungs of animals administered saline. Cytokine and chemokine profiles in bronchoalveolar lavage fluid indicated surface-chemistry dependence on modest secretion of IL-

6, IP-10, and MCP-1; however, there was no evidence of any deleterious histopathological changes. Based on these analyses, carbohydrate-functionalized nanoparticles are safe for in vivo applications. These results provide foundational information towards the evaluation of the capabilities of these surface-modified nanoparticles as vaccine delivery formulations.

5.2 Introduction

The development of novel strategies to improve adjuvant formulations by directly targeting innate immune cells is an important area of interest in the design of novel vaccines (1- 133

4). Biodegradable nanoparticles possess promising characteristics in this regard by playing dual roles, both as adjuvants and delivery vehicles (5). In particular, polyanhydride particles have been demonstrated to induce enhanced expression of MHC I and II on and stimulation of antigen presenting cells (APCs), which are fundamental to initiating adaptive immune responses (6-8).

After in vivo administration, the nanoparticles interact with a variety of cells, including APCs (9;

10). Before analyzing the effects of surface modification upon vaccine efficacy, an assessment of their safety and biocompatibility is necessary (11).

The use of polymeric nanoparticle systems for drug and vaccine delivery offers several advantages, including controlled delivery of encapsulated payload(s), and depending on their chemical properties, improved biocompatibility, receptor targeting capabilities, sustained antigen/drug release kinetics, adjuvanticity, and opportunities for both local and systemic delivery (12; 13). Polyanhydride nanoparticles have displayed these characteristics in both in vitro and/or in vivo settings (6; 9; 14-20). In particular, the use of biodegradable nanoparticles for lung delivery is an attractive proposition because of the following advantages: 1) uniform particle distribution in the lung; 2) local administration of vaccine antigens or therapeutic drugs;

3) sustained delivery of macromolecules; 4) improved patient compliance associated with non- invasive immunization and administration of fewer doses; and 5) avoidance of first pass metabolism, among others (2; 12; 21-23).

In previous in vitro studies, it was demonstrated that di-mannose functionalization of polyanhydride nanoparticles, which would induce signaling via C-type lectin receptors (CLRs) on APCs, enhanced the activation of macrophages and dendritic cells (DCs) (6; 17; 18; 24).

Because of the role of CLR signaling in stimulating innate immunity, identifying safe and effective means to selectively target APC receptors such as the macrophage mannose receptor

134

(MMR) and the macrophage galactose binding lectin (MGL) will provide novel approaches to enhance and shape adaptive immunity (3; 25; 26). Both the charge and surface properties of these polyanhydride nanoparticles are altered upon functionalization and could engage additional signaling cascade(s), which may affect the magnitude of immune response to the presence of these functionalized adjuvants/delivery vehicles. In this regard, even though these functionalized particles have displayed desirable properties (i.e., activation of APCs) in vitro, the focus of this study was to perform a systematic evaluation of their safety and biocompatibility profile in vivo to assess any toxicological effects that might be associated with functionalization.

5.3 Materials and Methods

5.3.1 Materials

Chemicals needed for monomer synthesis, polymerization, and nanoparticle synthesis included anhydrous (99+%) 1-methyl-2-pyrrolidinone (Aldrich, Milwaukee, WI); 1,6- dibromohexane, 4-p-hydroxybenzoic acid, N,N-dimethylacetamide and triethylene glycol (Sigma

Aldrich, St. Louis, MO); 4-p-fluorobenzonitrile (Apollo Scientific, Cheshire, UK); acetic acid, acetic anhydride, acetone, acetonitrile, dimethyl formamide (DMF), hexanes, methylene chloride, pentane, potassium carbonate, sodium hydroxide, sulfuric acid, and toluene (Fisher

Scientific, Fairlawn, NJ). For NMR characterization, deuterated dimethyl sulfoxide was purchased from Cambridge Isotope Laboratories (Andover, MA). For nanoparticle tracking

AlexaFluor® 647 hydrazide was purchased from Life Technologies (Grand Island, NY). For nanoparticle functionalization, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride,

N-hydroxysuccinimide, and ethylenediamine were purchased from Thermo Scientific (Waltham,

MA). Glycolic acid was purchased from Acros Organics (Pittsburgh, PA). Materials required for the lung tissue processing included: Dulbecco’s Modified Eagle Medium and Hank’s balanced

135

salt solution (Life Technologies, Grand Island, NY); HEPES buffer, penicillin-streptomycin, and

L-glutamine, (Mediatech, Herndon, VA); and heat inactivated fetal bovine serum (Atlanta

Biologicals, Atlanta, GA); and ammonium chloride, potassium bicarbonate, 0.5 M EDTA and sodium azide (Fisher Scientific). β-mercaptoethanol and rat immunoglobulin (rat IgG) were purchased from Sigma Aldrich (St. Louis, MO). Materials used for flow cytometry included: stabilizing cellular fixative solution (BD Bioscience, San Jose, CA); unlabeled anti-CD16/32

(i.e., anti-FcγR) (Southern Biotech, Birmingham, AL); FITC conjugated anti-mouse CD11c

(clone N418), PE conjugated anti-mouse CD11b (clone M1/70), Alexa Fluor® 700 conjugated anti-mouse F4/80 (clone BM8), PerCP/Cy5.5 conjugated anti-mouse Ly-6G/Ly-6C (Gr-1) (clone

RB6-8C5), PE/Cy7 anti-mouse CD326 (clone G8.8) and their corresponding isotype controls:

FITC conjugated Armenian Hamster IgG (clone HTK888), PE-conjugated rat IgG2bκ (clone

RTK4530), Alexa Fluor® 700 conjugated rat IgG2aκ (clone RTK2758), and PerCP/Cy5.5 conjugated rat IgG2bκ (clone RTK4530) and PE/Cy7 conjugated rat IgG2aκ (clone RTK2758)

(BioLegend, San Diego, CA). Endotoxin-free saline was obtained from the Iowa State University

College of Veterinary Medicine Pharmacy.

5.3.2 Monomer and polymer synthesis

Monomers of 1,6-bis(p-carboxyphenoxy) hexane (CPH) and 1,8-bis(p-carboxyphenoxy)-

3,6-dioxaoctane (CPTEG) were synthesized as described previously (15; 27). The 50:50

CPTEG:CPH copolymer was synthesized by melt polycondensation as previously described

(15). The chemical structure was characterized with 1H NMR using a Varian VXR 300 MHz spectrometer (Varian Inc., Palo Alto, CA). The synthesized 50:50 CPTEG:CPH copolymer had a

Mw of 10,500 Da, and the polydispersity index (PDI) of this copolymer was 1.5, which is consistent with previous work (15; 20).

136

5.3.3 Nanoparticle synthesis

Polyanhydride nanoparticles were synthesized using anti-solvent nanoencapsulation as described previously (28). Briefly, for flow cytometry, AlexaFluor® 647 hydrazide (1% w/w) and

20 mg/mL 50:50 CPTEG:CPH polymer were dissolved in methylene chloride (at 4°C). For histological and cytokine analyses, blank nanoparticles were synthesized. The polymer solution was sonicated at 40 Hz for 30 s using a probe sonicator (Ultra Sonic Processor VC 130PB,

Sonics Vibra Cell, Newtown, CT) and rapidly poured into a pentane bath (at -40°C) at a solvent to non-solvent ratio of 1:250. Particles were collected by filtration and dried under vacuum for

30 min.

5.3.4 Sugar synthesis

5.4.4.1 Synthesis of carboxymethyl α-1,2-linked dimannoside

Synthesis of carboxymethyl α-1,2-linked dimannose was carried out using fluorous-solid phase extraction (FSPE) as per literature procedure (29-31). Each glycosylation was performed with 2.0 equivalents of the donor in anhydrous dichloromethane at 0 ºC for 15 min. Facile purification of crude product by FSPE enabled easy preparation of the protected linear α-1,2- linked dimannose in high yield. FSPE was very helpful in the context of this particular synthesis as isolation of the target compound using regular silica gel chromatography turned out to be difficult owing to the formation of unwanted side products (hydrolyzed and rearranged donor).

The reducing terminal of the disaccharide was further functionalized by ozonolysis followed by

Jones oxidation to yield a carboxylic acid. Global deprotection was carried out using Birch reduction condition to produce the desired deprotected dimannoside (6).

137

5.4.4.2 Synthesis of carboxymethyl-β-galactoside

Allylated β-galactose acetate was subjected to Ruthenium catalyzed Sharpless oxidation, which resulted in carboxylic acid terminated β-galactose acetate in high yield. Base catalyzed deacetylation yielded the desired galactoside in high yield.

5.3.5 Surface functionalization carboxymethyl-α-D-mannopyranosyl-(1,2)-D-mannopyranoside and carboxymethyl- - galactoside were conjugated onto the surface of polyanhydride nanoparticles using an amine- carboxylic acid coupling reaction (6; 17; 18; 24). Particles with glycolic acid groups on the surface (linker) and non-functionalized (NF) particles were used as controls. The conjugation reaction was performed in two reaction steps, as described previously (6; 18). Briefly, a nanoparticle suspension (10 mg/mL) was made using nanopure water, and 10 equivalents of 1- ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and 12 equivalents of N- hydroxysuccinimide (NHS), and 10 equivalents of ethylenediamine were added. This reaction was carried out at 4°C temperature for 1 h at a constant agitation of 17 rcf. Following the reaction, the particles were centrifuged at 10,000 rcf for 10 min and the supernatant was removed. The particles were washed with the same volume of nanopure water, centrifuged at

10,000 rcf for 10 min and the supernatant was removed. A second reaction was performed with

10 eq. of EDC, 12 eq. of NHS and 10 eq. of the corresponding functionalizing agent (i.e., di- mannose or galactose) in nanopure water, using constant agitation at 17 rcf for 1 h at 4°C.

Particles were sonicated before and after each reaction to break aggregates. After the reactions were completed, nanoparticles were collected by centrifugation (10,000 rcf, 10 min) and dried under vacuum for 1 h.

138

5.3.6 Nanoparticle characterization

Morphological and size characterization of both the functionalized and the NF nanoparticles was performed using scanning electron microscopy (SEM, FEI Quanta 250, Kyoto,

Japan) and quasi-elastic light scattering (QELS, Zetasizer Nano, Malvern Instruments Ltd.,

Worchester, UK). The QELS experiments were used to measure the ζ-potential of the nanoparticles. To quantify the amounts of the carbohydrates conjugated to the nanoparticles, a high throughput version of a phenol-sulfuric acid assay was used (6; 32). A microplate reader

(SpectraMax M3, Molecular Devices, Sunnyvale, CA) was used to obtain the absorbance of standards and samples using a wavelength of 490 nm. The total amount of sugar per unit weight of nanoparticles (μg/mg) was calculated.

5.3.7 Mice

Female Swiss Webster outbred mice were purchased from Harlan Laboratories

(Indianapolis, IN). Mice were housed in specific pathogen-free conditions where all bedding, caging, and feed were sterilized prior to use. All animal procedures were conducted with the approval of the Iowa State University Institutional Animal Care and Use Committee.

5.3.8 Liver and kidney histological and biomarker examination

Separate groups of five Swiss Webster outbred mice were subcutaneously injected with 5 mg of polyanhydride nanoparticles (non- or surface-functionalized) in 1.5 mL of phosphate buffer saline (PBS) at the nape of the neck (33). Control animals received treatment that included

Alum (100 μL) or saline (1.5 mL). Urine samples were collected at 7 and 30 days post- administration, prior to necropsy. Whole blood was collected via cardiac puncture in heparinated tubes, and liver and kidney tissues were harvested during necropsy and placed in phosphate- buffered formalin. Formalin-fixed tissues after 7 and 30 days post-administration were

139

embedded, sectioned, and stained with hematoxylin and eosin (H&E), and blindly evaluated by a board-certified veterinary pathologist. Histopathological damage caused by inflammation, the distribution of inflammatory cells, and tissue necrosis were evaluated using a 0-5 scoring system for each independent parameter.

5.3.9 Serum biomarker analysis

Serum biomarkers of kidney and liver function were analyzed using an Ortho Vitros 5.1

Chemistry Analyzer by the Iowa State University Clinical Pathology Laboratory. Toxicological biomarkers analyzed included blood urea nitrogen (BUN), albumin, alkaline phosphatase (Alk

Phos), alanine aminotransferase (ALT), serum creatinine, glucose, total bilirubin, cholesterol and total triglycerides. Normal range values for these biomarkers were obtained from the Laboratory and compared with literature (34; 35).

5.3.10 Urine creatinine and total protein quantification analysis

Creatinine levels were measured in urine samples collected at 7 and 30 days post- administraton using a creatinine assay kit (Sigma Aldrich). The ELISA-based colorimetric assay was performed following the manufacturer’s specifications. Quantification of creatinine was performed using standards provided by the manufacturer. Total protein amount in urine was quantified using a micro-bicinchoninic acid (BCA) assay at an absorbance of 562 nm using a plate reader (SpectraMax M3).

5.3.11 Intranasal administration of particle formulations

Five separate groups of Swiss Webster mice were intranasally administered nanoparticle formulations. To sedate the mice prior to intranasal adminsitration of the nanoparticles, the mice were intraperitoneally injected with 90 μL of anesthetic solution (20 mg/mL ketamine + 1 mg/mL xylazine). Particle treatment groups included mice that were administered: 1) 500 μg of

140

non-functionalized, 2) linker-functionalized, 3) galactose-functionalized or 4) di-mannose functionalized 50:50 CPTEG:CPH nanoparticles. Mice intranasally administered saline were used as a control. Nanoparticles were suspended in PBS and sonicated before administration. For all formulations, a volume of 50 μL was intranasally administered. After mice were deeply anesthetized, they were held upright by the nape of the neck and nanoparticle suspension was slowly applied through the nostrils of each mouse with a micropipette. They were held in this position until the breathing rate of the animals was back to normal. Mice were monitored after anaesthesia and mobile function was restored.

5.3.12 Lung histological evaluation

Lungs from Swiss Webster mice were excised at 6 h, 24 h and 48 h post-immunization and formalin-fixed. Tissues were embedded, sectioned, and stained with H&E, and blindly evaluated by a board-certified veterinary pathologist. Adverse reactions in the lung tissue caused by inflammatory infiltration, necrosis, edema, bronchial associated lymphoid tissue (BALT) hyperplasia, and hemorrhage were evaluated using a 0-5 scoring system for each independent parameter.

5.3.13 Flow cytometric analysis of lung tissue

Mice were euthanized at 2 h, 24 h, and 48 h time points post-immunization. Lungs were processed as previously described (9). Briefly, lungs were excised and perfused with PBS. Lungs were incubated in Hank’s Balanced Salt Solution with 1 mg/mL collagenase D and 60 U/mL

DNAse II for 20 min at 37°C. Lung tissue was homogenized using a gentleMACS® tissue dissociator (Miltenyi Biotec, Cambridge, MA). To remove debris, samples were centrifuged for

250 rcf for 20 s, and filtered with a 40 μm cell filter. Red blood cells were lysed using ACK lysis buffer (150 mM ammonium chloride, 10 mM potassium bicarbonate, and 0.1 mM EDTA). Cell

141

samples (1 x 106 cells/mL) were blocked to prevent non-specific binding with 1% Rat IgG, 0.1% anti-mouse CD16/32 and 0.1% unconjugated Armenian Hamster IgG. Cells were surface stained with CD11c, CD11b, Ly6G/C Gr-1, and F4/80 markers. Samples were fixed with stabilizing fixative solution (BD Biosciences) and analyzed on a FACSCantoTM flow cytometer (Becton-

Dickinson, San Jose, CA) and the data was processed using FlowJo vX software (TreeStar Inc.,

Ashland, OR).

5.3.14 Bronchoalveolar lavage (BAL) fluid collection

BAL fluid was collected after 6 h, 24 h, and 48 h post immunization (36; 37). Briefly, after mice were euthanized, a sterile catheter was inserted into the exteriorized trachea of each mouse. Using a 1 mL syringe attached to the catheter, 1 mL of PBS was infused into the lungs, and aspirated back to the syringe. The process was repeated 2-3 times, per mouse, while massaging the chest externally. Samples were placed on ice, and centrifuged at 300 rcf for 30 s at

RT to remove cellular debris, and stored at -20°C until further analysis.

5.3.15 Cytokine and chemokine analysis

BAL fluid samples obtained at 6 h, 24 h and 48 h post-immunization were analyzed using a 13-plex cytokine and chemokine quantification kit (MILLIPLEX® MAP mouse cytokine/chemokine magnetic bead panel, EMD Millipore, Billerica, MA). Analytes quantified included: IL-6, KC, MCP-1, MIP-1α, MIP-2, IP-10, TNF-α, IFN-γ, IL-1β, IL-10, IL-12p40,

MIG, and RANTES. The assay was performed following manufacturer’s instructions, and data was acquired and analyzed using a Bio-Plex 200 TM system (Bio-Rad, Hercules, CA) as described in previous protocols (8; 17).

142

5.3.16 Statistical analysis

Statistical analysis was used to analyze the cell surface marker expression and cytokine secretion data. Two-way ANOVA and Dunnett’s test were used to determine statistical significance among treatments and p-values < 0.05 were considered significant.

5.4 Results

5.4.1 Functionalization and characterization of carbohydrate-modified nanoparticles

Our previous work has shown that amphiphilic nanoparticle chemistries were suitable for protein stabilization (16; 24; 28; 38; 39), demonstrated potent adjuvant responses (20), and were effectively taken up by and activated APCs (6; 17; 24; 40). Therefore, the 50:50 CPTEG:CPH nanoparticle formulation was chosen to perform the carbohydrate functionalization and to evaluate safety upon in vivo administration.

Particle morphology was characterized using scanning electron microscopy, as shown by the photomicrographs in Figure 5.1. The size of these particles was measured using ImageJ software (version 1.47v, NIH, Bethesda, MD). The diameter of the non-functionalized 50:50

CPTEG:CPH nanoparticles was 182 ± 59 nm. After functionalization, the diameter increased to

223 ± 61 nm, 228 ± 43 nm, and 236 ± 55 nm, for linker-, galactose-, and di-mannose-modified particles, respectively. In addition, the ζ-potentials and the surface concentration of the sugars for each formulation were measured and are shown in Figure 5.1. The NF particles are negatively charged, consistent with the presence of carboxylic acid moieties with a ζ-potential of -21 ± 3.2 mV, while the addition of the amine linker to which the neutral sugar moieties (i.e., galactose and di-mannose) were attached resulted in a positively charged surface with ζ-potentials of 18 ±

2.5 mV, 16 ± 2.1 mV and 19 ± 1.9 mV, respectively. After the carbohydrate modification was

143

completed, quantification of the amount of sugar linked to the particle surface was measured using a phenol sulfuric acid assay and indicated that 15 ± 5.5 μg of galactose or 19 ± 2.4 μg of di-mannose were present per milligram of particles.

5.4.2 Kidney histological evaluation and renal function

Following subcutaneous administration of 5 mg of particles to mice, serum, urine, and kidney samples were collected at 7 and 30 days. Histological evaluation of the tissue sections was performed and renal function biomarker levels were analyzed. Blood urea nitrogen (BUN) was measured in serum samples, while creatinine and total protein were quantified in urine samples. As shown in the representative histological images in Figure 5.2A, the inflammatory changes in the kidney during the period of the study were unremarkable as no significant differences were observed between the histological scores of mice treated with saline and the animals treated with the various nanoparticle formulations. Using a five point scale, the inflammatory infiltration scores ranged from 0-2, in all groups, with an average of 0.67, and these levels did not worsen between 7 to 30 days post-administration indicating that no acute or chronic inflammation was induced (Figure 5.2B). The distribution of the cellular infiltration had an average score of ~1.67, with only one mouse that was administered the galactose nanoparticles receiving a score of 3 at 7 days post-administration. In summary, these studies show that there was no histological evidence of tissue damage in the kidneys. Figure 5.2C shows renal function biomarker levels, BUN, and creatinine, both indicative of normal glomerular filtration rate. For the two biomarkers assessed and the total protein/creatinine ratio in the urine, there were no significant differences in urine samples collected at 7 or 30 days post- administration from mice that were treated with saline and samples from mice that received any of the nanoparticle treatments. Both BUN and creatinine levels were within previously reported

144

normal range levels (41; 42). Even though BUN levels were slightly lower than the reference values, there were no significant differences between the saline and particle groups. Variations in normal BUN levels have been previously reported to be mouse strain-dependent (34; 35; 43).

5.4.3 Liver histological evaluation and hepatic function

Seven and 30 days after subcutaneous administration of 5 mg of nanoparticle formulations, histological evaluation of the tissue sections was performed. Cholestasis and hepatocellular damage were evaluated in serum samples by measuring the levels of alkaline phosphatase (Alk Phos) and alanine aminotransferase (ALT), respectively. As shown in Figure

5.3A, the inflammatory changes in the liver were mild, and these were interpreted as non- specific background changes common to this mouse strain. Representative histological images of liver tissue are shown. The inflammatory infiltration scores ranged from 0-2 (on a scale of 0-5) in all the animals studied, with an average of ~1, regardless of the treatment as shown in Figure

5.3B. These scores did not significantly change between the two time points analyzed. The frequency of cellular infiltration within the liver (distribution score) was also low; the distribution score ranged from 0-3, with an average of ~1.8. In Figure 5.3C, the levels of Alk

Phos, ALT, and albumin are shown. There were no significant differences in the serum levels of any of the biomarkers analyzed between saline and nanoparticle treatment groups in the levels of these serum biomarkers at either time point. The ALT levels in mice administered the particle formulations were slightly higher than the upper limit of the reference values at day 30 post- administration; however, these were no different than the levels observed in mice treated with saline at the same time point.

145

5.4.4 Lung histological evaluation

Based upon the results obtained so far, which suggested that the subcutaneous administration of carbohydrate-functionalized polyanhydride nanoparticles did not have a detrimental effect upon liver or kidney function of the treated animals, we next evaluated the safety profile upon intranasal administration of surface-modified polyanhydride nanoparticles.

Previous work from our laboratories has shown that nanovaccines delivered intranasally resulted in protective long-term immunity (20). After intranasal administration of 0.5 mg of nanoparticle formulations, tissue samples were collected at 6, 24, and 48 h post-administration to evaluate acute histological changes. Figure 5.4A shows representative histological images from the lung for each treatment group at the 24 h time point (when the highest histological scores were registered). The parameter that contributed mostly to the final histological score was inflammation as shown in Figure 5.4B. Inflammatory infiltrates tended to be focused on bronchioles and adjacent alveolar spaces with neutrophils predominating. In general, the inflammatory scores peaked at 24 h, with scores up to 4 in some animals. The animals that received the linker and galactose-functionalized nanoparticles displayed a higher level of inflammation. In addition, the average necrosis values increased with time, with the highest value at 48 h, regardless of treatment groups (data not shown). There were only two mice with minor hemorrhage, with a score of 1, 24 h after administration of the NF particles. Hemorrhage was likely a tissue collection artifact. There were no mice with signs of edema or BALT hyperplasia at any time point analyzed (data not shown). Total histological scores are shown in

Figure 5.4C.

146

5.4.5 Distribution of lung cellular populations

Given the inflammatory cell infiltration scores described above, the cell types recruited into the lungs following particle administration were assessed by flow cytometry. Cellular populations were analyzed in whole lung homogenate at 2 h, 24 h, and 48 h post-intranasal administration. Flow cytometric analysis was performed and populations were identified with the following surface marker combinations: dendritic cells (CD11c+ CD11b-), interstitial macrophages (CD11c- CD11b+ F4/80+), neutrophils (CD11b+ Ly6G/C Gr-1+), and activated monocytes (CD11b+ Ly6G/C Gr-1-) (44-47). Figure 5.5 shows the cellular population distribution in the lungs. The percentage of DCs (Figure 5.5A) increased with time for all the treatment groups including the saline control and ranged from 1-4% of total lung cells. The percentage of interstitial macrophages (Figure 5.5B) peaked at 24 h, with all the treatment groups following similar dynamics. The neutrophil percentage (Figure 5.5C) at 2 h was the highest (2.5-

5%) and with time, these populations decreased to 1.5-2.5% of all the cells; this behavior was observed in the tissue from the animals that received all the treatment groups except the NF nanoparticles, in which the neutrophil population peaked at 24 h. As another measure of cellular recruitment into the lungs, the presence of activated monocytes was assessed (Figure 5.5D) and the presence of this cell type followed similar dynamics as neutrophils, starting at 4-6%, and decaying to 1.5-3% of total lung cells by 48 h. These data support the histological evaluation, as no major changes were observed in the inflammatory cell populations in the lungs of mice treated with saline or mice administered any of the nanoparticle formulations.

5.4.6 Cytokine/chemokine secretion

To assess the inflammatory environment in the lungs following intranasal administration of 0.5 mg of nanoparticle formulations, the BAL fluid was collected and used to measure the

147

amounts of the following chemokines and cytokines: IL-6, KC, MCP-1, MIP-1α, MIP-2, IP-10,

TNF-α, IFN-γ, IL-1β, IL-10, IL-12p40, MIG, and RANTES. Negligible amounts (i.e., below the levels of detection) of IFN-γ, IL-1β, IL-10, IL-12p40, MIG, and RANTES were observed (data not shown). Figure 5.6 shows the kinetics of the secretion of IL-6, KC, MIP-2, TNF-α, IP-10,

MCP-1, and MIP-1α. Two distinct trends were observed. The secretion of IL-6, KC, MIP-2, and

TNF-α peaked at 6 h post-administration, while the highest amounts of IP-10, MCP-1, and MIP-

1α secreted were observed 48 h post-administration. The BAL fluid from animals that received the linker and galactose-modified nanoparticle groups showed consistently higher amounts of these cytokines compared with saline. The absence of a major cytokine/chemokine response after intranasal administration of the nanoparticle formulations is consistent with the histological data and provides further evidence of the safety and biocompatibility of these materials following pulmonary delivery.

5.5 Discussion

In this work we report on the safety profile of surface-functionalized polyanhydride nanoparticles following parenteral or intranasal administration to mice. The safety and biocompatibility of non-functionalized polyanhydride nanoparticles has been demonstrated previously (33); however, the ability of di-mannose functionalized nanoparticles to initiate signaling via CLRs warrants a systematic evaluation of the potential toxicity associated with the induction of innate and/or inflammatory responses by these novel biomaterials.

As shown in Figure 5.1, the particle morphology and size of the functionalized particles were similar to previously reported data (6; 19; 28; 48; 49). The characterization of polyanhydride nanoparticles before and after surface modification was consistent with previous studies (6; 17; 18). The change in zeta potential from negative to positive charge after linker

148

attachment can be beneficial for enhanced cellular uptake, as reported previously (50; 51). In addition, carbohydrate functionalization is known to enhance both APC activation in vitro (6; 17;

18) as well as the therapeutic efficacy of drug delivery (4).

Based on the histological analysis scores reported in this work, subcutaneous administration of 5 mg of carbohydrate-functionalized nanoparticles did not result in tissue damage of the liver or kidney when assessed at 7 or 30 days post-administration and is consistent with previously reported safety profiles of parenterally administered polyanhydride nanoparticles

(33). The histological changes noted in hepatic tissue samples after nanoparticle administration were mild, which is comparable to other biodegradable polymer formulations, but much less than that exhibited by extremely cytotoxic metal particles of similar size (11). No other histological changes such as distortion and swelling of hepatocytes, cellular binucleation, or hydropic degeneration of the tissue were detected in the liver samples, as reported with other nanoparticle formulations (52). The histological scores of the kidney samples of mice administered saline alone were not significantly different when compared to the scores from the kidneys of mice administered the various particle formulations. No interstitial edema, inflammatory cell infiltration, tubular epithelial flattening, urinary casts, or signs of renal histopathological lesions were detected, which have been reported previously for other nanoparticles (53; 54). Based on the data presented here, the systemic effects of functionalized nanoparticle administration, in terms of hepatic and renal health, are consistent with effects of other biodegradable nanoparticle systems (33; 55; 56).

In addition to histological evaluation, serum and urine biomarkers, which are indicators of renal and hepatic injury and inflammation, were used to evaluate the safety of the functionalized nanoparticles. The levels of BUN, urine creatinine, and total protein/creatinine

149

ratio in mice receiving the particle treatments were not significantly different from the levels in animals receiving the saline. Even though BUN content level was lower than the reference values provided, the values were consistent with previously reported data in Swiss Webster mice

(43). Creatinine and total protein to creatinine ratio levels in mice administered the particles were no different than in animals receiving the saline control, and consistent with amounts reported in urine from mice of this lineage (35; 41). Creatinine, a byproduct of muscle metabolism, is an important indicator of kidney function. When glomerular filtration rate is impaired, creatinine levels rise in the blood and in the urine (57). Creatinine measurement is commonly used because urinary excretion of any biomarker that is filtered through the glomerulus is affected by the glomerular filtration rate and therefore used to normalize other markers, such as total protein or albumin (58). Together with urine specific gravity, these markers are part of the standard clinical diagnosis for renal function during impairments such as chronic kidney disease (59), acute kidney injury (60), or renal injury (41). Together, the inability to detect elevated levels of key biomarkers in serum and urine combined with the histopathology assessment of liver and kidney samples, demonstrates that there were no detrimental effects on renal or hepatic systems in mice treated with 5 mg of di-mannose functionalized nanoparticles.

The use of the pulmonary route offers several advantages for drug and vaccine delivery since the lungs allow targeted (2), non-invasive administration (22), and the capability to ensure systemic or local delivery of agents (12; 21). However, the respiratory system is also a more delicate environment, and parameters such as particle size (61-63), charge (2; 50; 51), chemistry

(9; 33), and material (12; 64) affect deposition, distribution, and biocompatibility. In previous work, a single intranasal administration of non-functionalized polyanhydride nanovaccines demonstrated the ability to induce protective immunity upon lethal challenge (9; 20; 65). The

150

current work builds upon these studies by evaluating the safety profile of carbohydrate- functionalized nanoparticles in the lung. The acute histopathology results from lung tissues after administration of particle formulations (Figure 5.4) displayed a bell-shaped curve with a peak at the 24 h time point. The lesions in the lung samples were mild to moderate, likely attributable to the administration procedure itself, because the lungs of the mice administered the saline control group received similar scores. As shown in Figure 5.4B, the major contributor to the histological scores was inflammatory cell infiltration.

The inflammatory infiltrates found in the lung samples after intranasal administration were consistent with the recruitment kinetics described for other pulmonary innate immune responses (36; 50; 61; 66), in which initial cellular recruitment was primarily composed of mononuclear cells and neutrophils. Next, neutrophils and macrophage infiltrates appeared in the lung tissue by 24 h, and were still present 48 h post-administration. Since the differences in the inflammatory cell infiltrates in the lung samples may be attributed to pulmonary recruitment of cells from circulation (50), we analyzed the kinetics of various cell populations in the lungs of treated animals. As shown in Figure 5.5, there were no significant changes in lung cellular populations between the various particle treatment groups, but all of them followed similar kinetics. Consistent with the histological data and previous studies (66; 67), neutrophils were the first cells to be recruited to the administration site, followed by macrophages and/or DCs.

Overall, the lung cellular populations observed were not statistically distinguishable from the populations in the lungs of mice receiving saline. This observation suggests that the administration of the nanoparticles likely caused a mild inflammatory response similar to that induced by the administration of saline, and supports the conclusion that the particle formulations themselves were not detrimental to the health of the treated animals.

151

Another component of lung response to foreign material is the secretion of cytokines and chemokines to recruit a cellular response and mediate clearance. The presence of these molecules can mediate leukocyte trafficking, inflammation, and link the innate and adaptive immune responses (68). However, overproduction of chemokines and cytokines can cause severe tissue damage (68; 69). As shown in Figure 5.6, different kinetics were observed for the analyzed cytokines and chemokines. The amounts of IL-6, KC, MIP-2 and TNF-α levels were elevated at early time points but decreased by 48 h, while the amounts of IP-10, MCP-1 and MIP-1α secreted were the highest at 48 h. The observed differential production of these cytokines/chemokines is likely related to the dynamic nature of the innate immune response and the different cell types that produce, utilize, and respond to these molecules. Nevertheless, our data indicate that intranasal administration of any of the nanoparticle formulations did not cause a major increase in cytokine or chemokine production that would result in severe tissue damage.

Altogether, the histological evaluation of lung tissue, the unremarkable changes in the recruitment of inflammatory cells to the lung, and the absence of a major cytokine/chemokine response after intranasal administration of carbohydrate-functionalized polyanhydride nanoparticles provide confirmatory evidence of the safety and biocompatibility of these novel materials for pulmonary delivery.

5.6 Conclusions

The studies reported herein demonstrate the safety and biocompatibility of carbohydrate- functionalized polyanhydride nanoparticles upon parenteral and intranasal administration. The results showed that a 5 mg dose of either linker- or di-mannose-functionalized nanoparticles did not induce hepatic or renal tissue damage or cause elevation of damage-related or functional biomarkers in serum or urine following subcutaneous administration. In addition, a 0.5 mg dose

152

of either linker- or di-mannose-functionalized nanoparticles administered intranasally did not result in demonstrable tissue changes in the lungs of treated animals. The favorable histological profile, distribution and kinetics of cellular populations, and the lack of a remarkable pro- inflammatory cytokine and chemokine profile in the lungs of mice administered functionalized nanoparticles supported the biocompatibility of the linker- and di-mannose-functionalized nanoparticles. Together, these studies demonstrate the safety of administering carbohydrate- functionalized nanoparticles in vivo and provide foundational information to evaluate the capabilities of these surface-modified nanoparticles for drug and vaccine delivery.

5.7 Acknowledgments

The authors would like to acknowledge financial support from NIH-NIAID (U19 AI-

091031) and U.S. Army Medical Research and Materiel Command (Grant No. W81XWH-10-1-

0806). The authors would also like to thank Shawn Rigby of the ISU Flow Cytometry Facility for his expertise in flow cytometry.

5.8 References 1. Engering A, Geijtenbeek TB, van Vliet SJ, Wijers M, van Liempt E, et al. 2002. The dendritic cell-specific adhesion receptor DC-SIGN internalizes antigen for presentation to T cells. Journal of immunology 168:2118-26

2. Azarmi S, Roa WH, Lobenberg R. 2008. Targeted delivery of nanoparticles for the treatment of lung diseases. Adv Drug Deliv Rev 60:863-75

3. Geijtenbeek TB, Gringhuis SI. 2009. Signalling through C-type lectin receptors: shaping immune responses. Nature reviews. Immunology 9:465-79

4. Irache JM, Salman HH, Gamazo C, Espuelas S. 2008. Mannose-targeted systems for the delivery of therapeutics. Expert Opin Drug Deliv 5:703-24

5. Ulery BD, Nair LS, Laurencin CT. 2011. Biomedical Applications of Biodegradable Polymers. J Polym Sci B Polym Phys 49:832-64

153

6. Carrillo-Conde B, Song EH, Chavez-Santoscoy A, Phanse Y, Ramer-Tait AE, et al. 2011. Mannose-functionalized "pathogen-like" polyanhydride nanoparticles target C-type lectin receptors on dendritic cells. Molecular pharmaceutics 8:1877-86

7. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

8. Petersen LK, Xue L, Wannemuehler MJ, Rajan K, Narasimhan B. 2009. The simultaneous effect of polymer chemistry and device geometry on the in vitro activation of murine dendritic cells. Biomaterials 30:5131-42

9. Ross KA, Haughney SL, Petersen LK, Boggiatto P, Wannemuehler MJ, Narasimhan B. 2014. Lung Deposition and Cellular Uptake Behavior of Pathogen-Mimicking Nanovaccines in the First 48 Hours. Advanced healthcare materials

10. Carrillo-Conde BR, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2012. Chemistry-dependent adsorption of serum proteins onto polyanhydride microparticles differentially influences dendritic cell uptake and activation. Acta biomaterialia 8:3618-28

11. Semete B, Booysen L, Lemmer Y, Kalombo L, Katata L, et al. 2010. In vivo evaluation of the biodistribution and safety of PLGA nanoparticles as drug delivery systems. Nanomedicine : nanotechnology, biology, and medicine 6:662-71

12. Mansour HM, Rhee YS, Wu X. 2009. Nanomedicine in pulmonary delivery. International journal of nanomedicine 4:299-319

13. Nguyen J, Steele TW, Merkel O, Reul R, Kissel T. 2008. Fast degrading polyesters as siRNA nano-carriers for pulmonary gene therapy. Journal of controlled release : official journal of the Controlled Release Society 132:243-51

14. Torres MP, Determan AS, Anderson GL, Mallapragada SK, Narasimhan B. 2007. Amphiphilic polyanhydrides for protein stabilization and release. Biomaterials 28:108-16

15. Torres MP, Vogel BM, Narasimhan B, Mallapragada SK. 2006. Synthesis and characterization of novel polyanhydrides with tailored erosion mechanisms. Journal of biomedical materials research. Part A 76:102-10

154

16. Carrillo-Conde B, Schiltz E, Yu J, Chris Minion F, Phillips GJ, et al. 2010. Encapsulation into amphiphilic polyanhydride microparticles stabilizes Yersinia pestis antigens. Acta biomaterialia 6:3110-9

17. Chavez-Santoscoy AV, Roychoudhury R, Pohl NL, Wannemuehler MJ, Narasimhan B, Ramer-Tait AE. 2012. Tailoring the immune response by targeting C-type lectin receptors on alveolar macrophages using "pathogen-like" amphiphilic polyanhydride nanoparticles. Biomaterials 33:4762-72

18. Vela Ramirez JE, Roychoudhury R, Habte HH, Cho MW, Pohl NL, Narasimhan B. 2014. Carbohydrate-functionalized nanovaccines preserve HIV-1 antigen stability and activate antigen presenting cells. Journal of biomaterials science. Polymer edition:1-20

19. Huntimer LM, Ross KA, Darling RJ, Winterwood NE, Boggiatto P, et al. 2014. Polyanhydride nanovaccine platform enhances antigen-specific cytotoxic T cell responses. Technology 02:171-5

20. Ulery BD, Kumar D, Ramer-Tait AE, Metzger DW, Wannemuehler MJ, Narasimhan B. 2011. Design of a protective single-dose intranasal nanoparticle-based vaccine platform for respiratory infectious diseases. PloS one 6:e17642

21. Patton JS, Byron PR. 2007. Inhaling medicines: delivering drugs to the body through the lungs. Nature reviews. Drug discovery 6:67-74

22. Sung JC, Pulliam BL, Edwards DA. 2007. Nanoparticles for drug delivery to the lungs. Trends in biotechnology 25:563-70

23. Bailey MM, Berkland CJ. 2009. Nanoparticle formulations in pulmonary drug delivery. Med Res Rev 29:196-212

24. Phanse Y, Carrillo-Conde BR, Ramer-Tait AE, Roychoudhury R, Pohl NL, et al. 2013. Functionalization of polyanhydride microparticles with di-mannose influences uptake by and intracellular fate within dendritic cells. Acta biomaterialia 9:8902-9

25. Napoletano C, Zizzari IG, Rughetti A, Rahimi H, Irimura T, et al. 2012. Targeting of macrophage galactose-type C-type lectin (MGL) induces DC signaling and activation. European journal of immunology 42:936-45

155

26. Cambi A, Figdor CG. 2003. Dual function of C-type lectin-like receptors in the immune system. Curr Opin Cell Biol 15:539-46

27. Conix A. 1966. Poly[1,3-bis(p-carxoyphenoxy)-propane anhydride]. Macro Synth 2:95-8

28. Ulery BD, Phanse Y, Sinha A, Wannemuehler MJ, Narasimhan B, Bellaire BH. 2009. Polymer chemistry influences monocytic uptake of polyanhydride nanospheres. Pharm Res 26:683-90

29. Curran DP, Luo Z. 1999. Fluorous Synthesis with Fewer Fluorines (Light Fluorous Synthesis): Separation of Tagged from Untagged Products by Solid-Phase Extraction with Fluorous Reverse-Phase Silica Gel. Journal of the American Chemical Society 121:9069-72

30. Zhang W, Curran DP. 2006. Synthetic applications of fluorous solid-phase extraction (F- SPE). Tetrahedron 62:11837-65

31. Zhang W. 2009. Fluorous linker-facilitated chemical synthesis. Chem Rev 109:749-95

32. Masuko T, Minami A, Iwasaki N, Majima T, Nishimura S, Lee YC. 2005. Carbohydrate analysis by a phenol-sulfuric acid method in microplate format. Analytical biochemistry 339:69-72

33. Huntimer L, Ramer-Tait AE, Petersen LK, Ross KA, Walz KA, et al. 2013. Evaluation of biocompatibility and administration site reactogenicity of polyanhydride-particle-based platform for vaccine delivery. Advanced healthcare materials 2:369-78

34. Mazzaccara C, Labruna G, Cito G, Scarfo M, De Felice M, et al. 2008. Age-Related Reference Intervals of the Main Biochemical and Hematological Parameters in C57BL/6J, 129SV/EV and C3H/HeJ Mouse Strains. PloS one 3:e3772

35. Cheetham SA, Smith AL, Armstrong SD, Beynon RJ, Hurst JL. 2009. Limited variation in the major urinary proteins of laboratory mice. Physiology & behavior 96:253-61

36. Kirby AC, Raynes JG, Kaye PM. 2006. CD11b regulates recruitment of alveolar macrophages but not pulmonary dendritic cells after pneumococcal challenge. The Journal of infectious diseases 193:205-13

156

37. Kirby AC, Coles MC, Kaye PM. 2009. Alveolar macrophages transport pathogens to lung draining lymph nodes. Journal of immunology 183:1983-9

38. Kipper MJ, Shen E, Determan A, Narasimhan B. 2002. Design of an injectable system based on bioerodible polyanhydride microspheres for sustained drug delivery. Biomaterials 23:4405-12

39. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

40. Petersen LK, Ramer-Tait AE, Broderick SR, Kong CS, Ulery BD, et al. 2011. Activation of innate immune responses in a pathogen-mimicking manner by amphiphilic polyanhydride nanoparticle adjuvants. Biomaterials 32:6815-22

41. O'Bryan T, Weiher H, Rennke HG, Kren S, Hostetter TH. 2000. Course of renal injury in the Mpv17-deficient transgenic mouse. Journal of the American Society of Nephrology : JASN 11:1067-74

42. Dunn SR, Qi Z, Bottinger EP, Breyer MD, Sharma K. 2004. Utility of endogenous creatinine clearance as a measure of renal function in mice. Kidney international 65:1959-67

43. Eaton KA, Friedman DI, Francis GJ, Tyler JS, Young VB, et al. 2008. Pathogenesis of renal disease due to enterohemorrhagic Escherichia coli in germ-free mice. Infection and immunity 76:3054-63

44. Lee PY, Wang JX, Parisini E, Dascher CC, Nigrovic PA. 2013. Ly6 family proteins in neutrophil biology. Journal of leukocyte biology 94:585-94

45. Lai L, Alaverdi N, Maltais L, Morse HC, 3rd. 1998. Mouse cell surface antigens: nomenclature and immunophenotyping. Journal of immunology 160:3861-8

46. Vermaelen K, Pauwels R. 2005. Pulmonary dendritic cells. American journal of respiratory and critical care medicine 172:530-51

47. Lohmann-Matthes ML, Steinmuller C, Franke-Ullmann G. 1994. Pulmonary macrophages. The European respiratory journal 7:1678-89

157

48. Petersen LK, Phanse Y, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2012. Amphiphilic polyanhydride nanoparticles stabilize Bacillus anthracis protective antigen. Molecular pharmaceutics 9:874-82

49. Haughney SL, Petersen LK, Schoofs AD, Ramer-Tait AE, King JD, et al. 2013. Retention of structure, antigenicity, and biological function of pneumococcal surface protein A (PspA) released from polyanhydride nanoparticles. Acta biomaterialia 9:8262-71

50. Harush-Frenkel O, Bivas-Benita M, Nassar T, Springer C, Sherman Y, et al. 2010. A safety and tolerability study of differently-charged nanoparticles for local pulmonary drug delivery. Toxicology and applied pharmacology 246:83-90

51. Nguyen J, Reul R, Betz T, Dayyoub E, Schmehl T, et al. 2009. Nanocomposites of lung surfactant and biodegradable cationic nanoparticles improve transfection efficiency to lung cells. Journal of controlled release : official journal of the Controlled Release Society 140:47- 54

52. Alarifi S, Ali D, Al-Doaiss AA, Ali BA, Ahmed M, Al-Khedhairy AA. 2013. Histologic and apoptotic changes induced by titanium dioxide nanoparticles in the livers of rats. International journal of nanomedicine 8:3937-43

53. Yamagishi Y, Watari A, Hayata Y, Li X, Kondoh M, et al. 2013. Acute and chronic nephrotoxicity of platinum nanoparticles in mice. Nanoscale research letters 8:395

54. Kelley VE, Winkelstein A. 1980. Age- and sex-related glomerulonephritis in New Zealand white mice. Clinical immunology and immunopathology 16:142-50

55. Yang Y, Pan D, Luo K, Li L, Gu Z. 2013. Biodegradable and amphiphilic block copolymer-doxorubicin conjugate as polymeric nanoscale drug delivery vehicle for breast cancer therapy. Biomaterials 34:8430-43

56. Italia JL, Bhatt DK, Bhardwaj V, Tikoo K, Kumar MN. 2007. PLGA nanoparticles for oral delivery of cyclosporine: nephrotoxicity and pharmacokinetic studies in comparison to Sandimmune Neoral. Journal of controlled release : official journal of the Controlled Release Society 119:197-206

57. Zhang XD, Wu D, Shen X, Liu PX, Fan FY, Fan SJ. 2012. In vivo renal clearance, biodistribution, toxicity of gold nanoclusters. Biomaterials 33:4628-38

158

58. Goldstein SL. 2010. Urinary kidney injury biomarkers and urine creatinine normalization: a false premise or not? Kidney international 78:433-5

59. Peralta CA, Shlipak MG, Judd S, Cushman M, McClellan W, et al. 2011. Detection of chronic kidney disease with creatinine, cystatin C, and urine albumin-to-creatinine ratio and association with progression to end-stage renal disease and mortality. JAMA : the journal of the American Medical Association 305:1545-52

60. Waikar SS, Sabbisetti VS, Bonventre JV. 2010. Normalization of urinary biomarkers to creatinine during changes in glomerular filtration rate. Kidney international 78:486-94

61. Nemmar A, Hoylaerts MF, Hoet PH, Vermylen J, Nemery B. 2003. Size effect of intratracheally instilled particles on pulmonary inflammation and vascular thrombosis. Toxicology and applied pharmacology 186:38-45

62. Heyder J, Gebhart J, Rudolf G, Schiller CF, Stahlhofen W. 1986. Deposition of particles in the human respiratory tract in the size range 0.005–15 μm. J Aerosol Sci 17:811-25

63. Hofmann W. 2011. Modelling inhaled particle deposition in the human lung—A review. J Aerosol Sci 42:693-724

64. Choi HS, Ashitate Y, Lee JH, Kim SH, Matsui A, et al. 2010. Rapid translocation of nanoparticles from the lung airspaces to the body. Nature biotechnology 28:1300-3

65. Ulery BD, Petersen LK, Phanse Y, Kong CS, Broderick SR, et al. 2011. Rational design of pathogen-mimicking amphiphilic materials as nanoadjuvants. Scientific reports 1:198

66. Dailey LA, Jekel N, Fink L, Gessler T, Schmehl T, et al. 2006. Investigation of the proinflammatory potential of biodegradable nanoparticle drug delivery systems in the lung. Toxicology and applied pharmacology 215:100-8

67. Xing Z, Gauldie J, Cox G, Baumann H, Jordana M, et al. 1998. IL-6 is an antiinflammatory cytokine required for controlling local or systemic acute inflammatory responses. The Journal of clinical investigation 101:311-20

68. Coelho AL, Hogaboam CM, Kunkel SL. 2005. Chemokines provide the sustained inflammatory bridge between innate and acquired immunity. Cytokine & growth factor reviews 16:553-60

159

69. Frevert CW, Huang S, Danaee H, Paulauskis JD, Kobzik L. 1995. Functional characterization of the rat chemokine KC and its importance in neutrophil recruitment in a rat model of pulmonary inflammation. Journal of immunology 154:335-44

5.9 Figures

Figure 5.1 Polyanhydride nanoparticle characterization. Chemical structures of the surface moieties on non-functionalized, linker-, galactose-, and di-mannose-functionalized nanoparticles are presented. Particle size data represent the mean ± standard deviation (SD) of data collected from scanning electron microscopy photomicrographs using ImageJ software from four independent experiments. Zeta potential data were measuring using QELS and represent the

160

mean ± SD of four independent experiments. Sugar density data were measured using a phenol sulphuric acid assay and are presented as the mean ± SD of four independent experiments.

Figure 5.2 Administration of a 5 mg dose of surface-functionalized 50:50 CPTEG:CPH nanoparticles did not affect renal function. Panel A shows representative histological sections of kidney samples from Swiss Webster mice (n= 6) seven days post-administration of the nanoparticle formulations. Panel B shows the inflammatory scores of kidney samples after histopathological evaluation. Panel C displays the levels of blood urea nitrogen (BUN) in serum and creatinine and total protein/creatinine ratio in urine samples from Swiss Webster mice seven and 30 days post-administration. Reference levels provided by the Iowa State University Clinical

161

Pathology Laboratory are indicated as dashed lines. No significant differences were observed when compared to mice administered saline (n = 5 at each time point).

Figure 5.3 Administration of a 5 mg dose of surface-functionalized 50:50 CPTEG:CPH nanoparticles did not affect induce hepatic inflammation or alter hepatic function. Panel A shows representative histological sections of liver samples from Swiss Webster mice (n= 6) seven days post-administration of the nanoparticle formulations. Panel B shows the inflammatory scores of liver samples after histopathological evaluation. Panel C displays the levels of alkaline phosphatase, alanine aminotransferase, and albumin in serum samples from Swiss Webster mice seven and 30 days post-administration. Reference levels provided by the Iowa State University

Clinical Pathology Laboratory are indicated as dashed lines. No significant differences were observed when compared to mice administered saline (n = 5 at each time point).

162

Figure 5.4 Mild to moderate inflammation was observed in lung samples upon histological evaluation post-administration of surface-functionalized polyanhydride nanoparticles. Lung tissue samples from Swiss Webster mice were collected at 6, 24 and 48 hours post- administration. A) Representative images of each treatment group at 24 h post-administration. B)

Inflammatory infiltration scores on a scale of 0-5 after histopathological evaluation. C)

Composite histopathological scores representing the sum of five individual parameters

(inflammatory infiltration, necrosis, edema, bronchial associated lymphoid tissue hyperplasia, and hemorrhage), with a total possible score of 25. No significant differences were observed when compared to mice administered saline (n = 6 at each time point).

163

Figure 5.5 Cellular distribution of lung homogenates of Swiss Webster following administration of surface-functionalized polyanhydride nanoparticles. Cellular populations were analyzed by flow cytometry at 6, 24 and 48 hours post-administration and various cell populations were analyzed. A) dendritic cells; B) interstitial macrophages; C) neutrophils; and D) activated monocytes. No significant differences were observed in these distributions when compared to mice administered saline (n = 6 at each time point).

164

Figure 5.6 Low amounts of cytokine/chemokine secretion were observed in bronchoalveolar lavage fluid 6, 24 and 48 hours after intranasal administration of surface-functionalized polyanhydride nanoparticles. The amounts of IL-6, KC, MIP-2, TNF-α, IP-10, MCP-1 and MIP-

1α in the bronchoalveolar lavage fluid were quantified using a Multiplex magnetic bead assay. * represents groups that are statistically significant (p ≤ 0.05) compared to the saline control (n = 5 at each time point).

165

CHAPTER 6: CELL- AND ANTIBODY-MEDIATED IMMUNITY INDUCED BY PATHOGEN-MIMICKING NANOVACCINES

Jonathan T. Goodman1, Danielle A. Wagner2, Paola M. Boggiatto2, Rajarshi Roychoudhury3,

Nicola L. B. Pohl3, Michael J. Wannemuehler2, Balaji Narasimhan1

To be Submitted to Acta Biomaterialia December, 2015

1 Department of Chemical and Biological Engineering, Iowa State University, Ames, IA 50011, U.S.A. 2 Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, Ames, IA 50011, U.S.A. 3 Department of Chemistry, Indiana University, Bloomington, IN 47405, U.S.A.

166

6.1 Abstract Enhancing the activation of antigen presenting cells (APCs) by receptor-mediated endocytosis of nanoparticles is a promising approach to tailoring the adaptive immune response.

C-type lectin receptors (CLRs) are a class of pattern recognition receptors (PRRs) that bind to carbohydrates and that are expressed on APCs. In this work, the carbohydrate di-mannose was covalently functionalized to the surface of polyanhydride nanoparticles in order to interact with the macrophage mannose receptor (MMR). These nanoparticles contained ovalbumin (Ova) and were administered subcutaneously to C57BL/6 mice. Using an adoptive transfer model, the antigen-specific T cell response induced by these formulations was evaluated. The non- functionalized nanoparticles and soluble Ova alone induced the greatest CD8+ T cell and induced strong activation of the transgenic CD8+ T cells as evidenced by the population of OTI cells containing intracellular IFN-γ and TNF-α. In contrast, the carbohydrate-functionalized nanoparticles induced less antigen-specific CD8+ T cell expansion. A memory recall response was observed in the animals that received non-functionalized or soluble protein formulations when the mice were challenged with Ova-expressing tumor cells. In the weeks following primary immunization, animals receiving the functionalized or non-functionalized nanoparticles showed significantly higher antibody titers compared to that induced in animals receiving soluble protein. It was shown that the surface functionalization of the nanoparticles is labile and is lost within a few hours of incubation in aqueous solutions. These results provide important insights into the complex interactions between nanoparticle chemistry and innate and adaptive immunity.

6.2 Introduction The induction of long lasting and protective immune responses is an important deliverable of many vaccine regimens. In this context, the inability to induce cellular responses has led to a lack of efficacious vaccines for diseases requiring avid T cell responses, such as HIV

167

or tuberculosis (1; 2). One strategy to overcome this hurdle is to enhance the activation of antigen presenting cells (APCs), such as dendritic cells (DCs), by exploiting receptor-mediated uptake of vaccine formulations. With a wide array of pattern recognition receptors (PRRs) to internalize antigen and activate downstream signaling, DCs lie at the crossroads of the innate and adaptive immunity (3). They efficiently present antigen to cognate T cells and provide important co-stimulatory signals that direct the outcome of the immune response

An important class of PRRs expressed on DCs are C-type lectin receptors (CLRs) (4).

There are over 60 types of CLRs expressed on the surface of APCs and they recognize a diverse range of carbohydrate structures (4). Some well characterized CLRs include cluster of differentiation (CD) 205 (also referred to as DEC-205), dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN), and the macrophage mannose receptor

(MMR). Engaging the MMR on DC and macrophage (MØ) cultures in vitro has been the focus of recent efforts from our laboratories (5-8). In these studies, polyanhydride microparticles and nanoparticles functionalized with di-mannose were internalized more efficiently by phagocytic cells. Additionally, incubating these carbohydrate-functionalized particles to DC or MØ cultures resulted in the secretion of higher amounts of pro-inflammatory cytokines (e.g., TNF, IL-6) and upregulated cell surface marker expression (e.g., CD80/86, CD40, MHC II) on these cells relative to that induced by non-functionalized particles.

In our previous work, we showed that polyanhydride nanovaccines induced long-lived antibody responses for at least 40 weeks post-immunization (PI) (9). The immunomodulatory capabilities of these nanovaccines helped enhance isotype switching and promoted higher antibody titers compared to a regimen consisting of soluble antigen (10; 11). Additionally, these vaccine formulations induced cell-mediated immunity more effectively than alum and also

168

elicited robust recall responses (10). The nanoparticles also provide prolonged release of a variety of encapsulated antigens that maintain both conformational and biological activity upon release (12-14). Formulations consisting of a soluble bolus with encapsulated antigen were shown to induce the most robust adaptive immune responses compared to an regimen on which all of the antigen was encapsulated within particles, priming the immune response against the antigen, and providing a prolonged presence of antigen to enhance the development of B cells

(11; 15; 16).

Towards rational design of vaccines, this work focused on monitoring the adaptive immune response induced by di-mannose functionalized polyanhydride nanovaccines, using ovalbumin (Ova) as a model antigen. Here, CD8+ and CD4+ OVA-specific, transgenic T cells were adoptively transferred to C57BL/6 mice to evaluate the antigen-specific T cell response induced by a single dose of carbohydrate functionalized nanoparticles. In addition, serum was collected from the animals and analyzed for anti-Ova IgG titers. Our results indicate that carbohydrate-functionalized nanoparticles did not outperform non-functionalized nanoparticles in inducing robust expansion of the Ova-specific T cells or avid antibody responses.

6.3 Materials and Methods

6.3.1 Materials For CPH and CPTEG monomer synthesis 1,6-dibromohexane, triethylene glycol, 4-p- hydroxybenzoic acid, and 1-methyl-2-pyrrolidinone were purchased from Sigma-Aldrich (St.

Louis, MO). 4-p-fluorobenzonitrile was obtained from Apollo Scientific (Cheshire, UK), while toluene, sulfuric acid, acetonitrile, dimethyl formamide, acetic anhydride, methylene chloride, pentane, and potassium carbonate were obtained from Fisher Scientific (Fairlawn, NJ).

Polymerization of the monomers required acetic anhydride, methylene chloride, and hexane

169

which were also purchased from Fisher Scientific. 1H NMR analysis was conducted using deuterated chloroform from Cambridge Isotope Laboratories (Andover, MA). Chemicals from

Fisher Scientific were used in the synthesis of nanoparticles including methylene chloride, and pentane. Ovalbumin was purchased from Sigma-Aldrich and passed through a Detoxi-Gel

Endotoxin removing column from Thermo Scientific. Materials for nanoparticle functionalization included N-hydroxysuccinimide (NHS) from Fisher Scientific, ethylenediamine from Sigma Aldrich, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and glycolic acid from Acros Organics (Geel, Belgium).

Adoptive transfer experiments required sterile PBS from Mediatech (Manassas, VA) and heparin from Fresenius Kabi USA (Waltham, MA). Complete tissue culture medium (CTCM) was composed of β-mercaptoethanol (Sigma-Aldrich), heat inactivated fetal bovine serum

(Atlanta Biologicals, Atlanta, GA), and RPMI 1640, penicillin/streptomycin, and L-glutamine, all purchased from Mediatech. Ammonium-Chloride-Potassium (ACK) lysing buffer consisted of ammonium chloride, potassium bicarbonate, and ethylenediaminetetraacetic acid (EDTA), all of which were purchased from Fisher Scientific. Cells were stimulated in phorbol 12-myristate

13-acetate (PMA), ionomycin, and brefeldin A, all of which were acquired from Sigma-Aldrich.

Flow cytometry was performed using the Fix/Perm Buffer kit from eBioscience. FITC-labeled anti-mouse Thy1.2 (clone 53-2.1), PE-labeled anti-mouse TNF-α (clone MP6-XT22), APC- labeled anti-mouse CD8b (clone eBioH35-17.2), APC-labeled anti-mouse CD4 (clone GK1.5),

PerCp-Cy5.5-labeled anti-mouse IL-2 (clone JES-5H4) and PE-Cy7-labeled anti-mouse IFN-γ

(clone XMG1.2) were obtained from Biolegend or eBioscience. Compensation beads were obtained from eBioscience. Cells were washed in FACS buffer which consisted of PBS

(pH=7.4), sodium azide from Fisher Scientific, and 2% heat inactivated fetal bovine serum.

170

Enzyme-linked immunosorbent assays (ELISAs) were completed with AP-conjugated anti-mouse IgG (H&L), AP-conjugated anti-mouse IgG2c, or AP-conjugated anti-mouse IgG1, all of which were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA),

PBS-Tween (0.2% Tween), and p-nitrophenyl phosphate (Fisher Scientific). Wells were blocked with 2 wt. % gelatin from Becton Dickinson (Sparks, MD).

6.3.2 Monomer, polymer, and nanoparticle synthesis 1,6-bis(p-carboxyphenoxy)hexane (CPH) and 1,8-bis(p-carboxyphenoxy)-3,6- dioxaoctane (CPTEG) diacids were synthesized as described previously (17; 18). These diacids were used in a melt polycondensation reaction to synthesize 20:80 CPTEG:CPH copolymer

(Figure 6.1A) (19). Purity, chemical structure, and molecular weight of the polymers were determined using 1H NMR spectroscopy on a Varian VXR 300 MHz spectrometer. Differential scanning calorimetry (DSC, TA Instruments) and gel permeation chromatography (GPC, Waters) were used to determine polymer glass transition temperature and polydispersity, respectively.

The 20:80 CPTEG:CPH copolymer had a molecular weight of 9.2 kDa from 1H NMR measurements, which is consistent with previous work (13; 20). The actual molar ratio was determined to be 22:78 CPTEG:CPH using 1H NMR analysis.

Polyanhydride nanoparticles were synthesized using an anti-solvent nanoencapsulation technique (21). A 20 mg/mL polymer solution with 2% (w/w) Ova was dispersed in methylene chloride and poured into a pentane bath so that the ratio of solvent to anti-solvent was 1:250.

This procedure was carried out at 4 °C with the pentane bath held at -40 °C and the particles were collected by vacuum filtration. Particle morphology was examined by a FEI Quanta 250 scanning electron microscope (SEM, Kyoto, Japan) after being coated with 2 nm of iridium using a Quorum Q150TS sputter coater (Lewes, UK). Nanoparticle size distribution was

171

characterized using ImageJ image analysis software (National Institutes of Health, Bethesda,

MD) with an average of 200 nanoparticles per image.

6.3.3 Nanoparticle functionalization The particles were functionalized with either di-mannose or a glycolic acid linker, which links the di-mannose to the ethylenediamine. Fluorous solid-phase extraction-based iterative synthesis was performed to obtain α-1,2-linked-di-mannopyranoside with a fluorous allyl group

(22). The fluorous tag was cleaved by ozonolysis, while oxidation with Jones Reagent produced a carboxylic acid group for coupling to nanoparticles. Birch reduction was used to yield a deprotected carboxymethyl-modified disaccharide of 1,2-α-linked-D-mannopyranoside (23). A previously optimized two-step amide coupling reaction covalently linked the di-mannose to the nanoparticles (6; 7). The first reaction involved incubating the nanoparticles with 10 molar equivalents of EDC, 12 molar equivalents of NHS, and 10 molar equivalents of ethylenediamine for one hour on a tube rotor at 4 °C. The nanoparticles were then washed with nanopure water and were collected by centrifugation at 10,000 rcf and 10 min. After being sonicated briefly at 40

Hz, the nanoparticles were incubated with 10 molar equivalents of EDC, 12 molar equivalents of

NHS, and 10 molar equivalents of di-mannose for one hour at 4 °C allowing the di-mannose to attach to the free amine of the ethylenediamine on the surface of the nanoparticles (Figure 6.1B).

6.3.4 Characterization of functionalized nanoparticles The sugar density on the surface of the nanoparticles was quantified using a phenol- sulfuric acid assay performed in flat bottom 96-well plates (24). A reaction time of 30 min at 95

°C was used for the sugars on the nanoparticle surface to be reduced inducing a color change in the solution. The absorbance of each well was measured at a wavelength of 490 nm using spectrophotometry (SpectraMax M3, Molecular Devices, Sunnyvale, CA). This was compared against soluble di-mannose standards, while non-functionalized nanoparticles were used as a

172

negative control. Energy dispersive X-ray spectroscopy (EDS) was also performed to assay for nitrogen content from the ethylenediamine on the particles.

To track the zeta potential of the non-functionalized and functionalized nanoparticles as a function of time, 5 mg of nanoparticles and 1 mL PBS were placed in 1.5 mL microcentrifuge tube, sonicated, and incubated at 37°C in a 100 rpm rotator. 40 μL of particle suspension was removed at each time point and diluted in 960 μL of cold nanopure water for a final concentration of 200 μg/mL. The surface potential was measured in mV using a ZetaSizer Nano

(Worchester, UK).

6.3.5 Animals Recipient female C57BL/6 Thy1.1, donor OT-I Thy1.2 C57BL/6 mice, and donor OT-II

Thy1.2 C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, ME). All mice were housed under specific pathogen-free condition with all the bedding, caging, water, and feed sterilized prior to use. The Iowa State University Institutional Animal Care and Use Committee approved all animal procedures.

6.3.6 Adoptive transfer and mouse immunizations 7.5 x 105 OT-I Thy1.2+ CD8+ T cells were transferred into Thy1.1+ recipient C57BL/6 mice on Day -1. Lymph nodes of the OT-I donors were harvested, processed, and the number of cells were counted using a Beckman Coulter Z1 particle counter. Flow cytometry was performed on an aliquot of the cells to assess the frequency of Thy1.2+ CD8+ T cells in order to calculate the total number of OT-I cells to be administered. Adoptive transfers were performed via a tail vein injection of 100 μL of OT-I cells in PBS. One day later the recipient mice were immunized subcutaneously at the nape of the neck with 200 μL of PBS containing one of the vaccine treatments. There were five immunization groups each consisting of eight mice with one group

173

serving as a naïve (i.e., saline-administered) control and another group receiving 100 μg of soluble Ova, as previously described (25). Separate groups of mice also received either non- functionalized Ova-containing 20:80 CPTEG:CPH nanoparticles, Ova-containing linker- functionalized 20:80 CPTEG:CPH nanoparticles, or di-mannose functionalized 20:80

CPTEG:CPH nanoparticles. In the particle formulations, 90 μg of the Ova was administered as soluble antigen and 10 μg were encapsulated into 500 μg of nanoparticles (i.e, 2 % loaded). The same experimental design was utilized to evaluate the Thy1.2+ CD4+ T cell responses.

6.3.7 T cell expansion and intracellular cytokine staining At days 2, 6, and 10 PI, approximately 100 μL of blood was collected from the saphenous vein of each mouse and dispensed into PBS containing 500 unit/mL heparin solution to prevent clotting. ACK lysis buffer was added to the suspension to lyse red blood cells. CTCM medium was added following 3 min of lysing to neutralize the lysis buffer. The cells were centrifuged at

250 xg at 4 °C for 10 min and the lysing step was repeated once more. Cells were enumerated and 5 x 105 total cells were added to each well of a 96 well microtiter plate. These cells were stimulated at 37 °C in a 5% CO2 environment for 5 h with CTCM medium that consisted of 50 ng/mL PMA, 500 ng/mL ionomycin, and 10 μg/mL brefeldin A. Cells were also stimulated with brefeldin A only to serve as a baseline control. Fixative/permeation buffer was added for overnight incubation at 4°C and the following day the cells were labeled with monoclonal antibodies directed at surface markers and for detection of intracellular cytokines. The cells were incubated in FACS buffer and were analyzed using a BD FACS Canto flow cytometer.

6.3.8 Antigenic challenge 5 x 106 EG7 ovalbumin expressing tumor cells were administered to the right rear flank four weeks PI to OT-I and OT-II recipient mice. Four days post challenge, 100 μL of peripheral

174

blood was collected from each mouse and the intracellular cytokine staining procedure was repeated to determine the efficacy of a memory response.

6.3.9 Antibody titer Flat-bottomed, high-binding 96 well plates were coated with 100 μL of 5 μg/mL Ova in each well and allowed to incubate at 4°C overnight. On the following day, the wells were emptied and 2% (w/v) gelatin blocking buffer was allowed to incubate for 2 h at room temperature. The blocking buffer was removed and the plates were washed in PBS-Tween

(PBST) buffer. Serum was added to the wells and 1:2 dilutions were performed across the plate in antibody dilution buffer consisting of PBS and 1% (v/v) normal goat serum. The plates were allowed to incubate at 4°C overnight. On the third day, the plates were washed in PBST and alkaline phosphatase-conjugated goat anti-mouse IgG (or a subclass of IgG) was added at 1:1000 dilution. After 2 h of incubation at room temperature, the plates were washed in PBST and 100

µL (1 mg/mL) of alkaline phosphatase substrate was added. One hour later, the plates were analyzed at a wavelength of 405 nm using a spectrophotometer.

6.3.10 Statistical analysis An unpaired t test was performed to determine statistical significance among the treatments. A p-value less than 0.05 was considered significant.

6.4 Results

6.4.1 Nanoparticle characterization Nanoparticles based on 20:80 CPTEG:CPH were synthesized for use in this work because this chemistry has previously been demonstrated to induce robust antibody titers and induce a cell-mediated response (Figure 6.1) (9; 25). When added to DC cultures in vitro, carbohydrate modification of 20:80 CPTEG:CPH nanoparticles resulted in enhanced cell surface marker expression and induced secretion of pro-inflammatory cytokines by the DCs compared to

175

the addition of nanoparticles of other chemistries (7). Using SEM, the nanoparticle size was determined to be approximately 180 nm and the morphology was comparable to that observed in previous studies (Table 6.1) (7; 9). Carbohydrate functionalization did not significantly affect nanoparticle size. Carbohydrate modification was characterized by EDS and the colorimetric phenol-sulfuric acid assay. A strong nitrogen peak was detected on both the linker and di- mannose functionalized nanoparticles confirming the presence of the ethylenediamine (Table

6.1). The phenol-sulfuric acid assay indicated the presence of di-mannose on the nanoparticles relative to the non-functionalized nanoparticles. The results of the EDS and phenol-sulfuric acid assay were also consistent with previous work (5-7; 26).

The zeta potential was recorded for all the nanoparticle formulations as a function of time during incubation in PBS (Figure 6.2). Upon suspension in PBS, the zeta potential of the negatively charged non-functionalized particles was -19 mV and is consistent with previous observations (7; 12; 26). In contrast, the initial zeta potential of the linker-functionalized particles was +13 mV, as expected, because of the presence of the positively charged amine groups. When these measurements were carried out as a function of time, it was observed that the surface charge of the linker-functionalized particles decreased. After 2-3 h of incubation, the surface charge became negative and plateaued after 24 hours at approximately -10 mV. In contrast, the zeta potential of the non-functionalized nanoparticles remained unchanged for the entire duration of the experiment. These results indicate that the engagement of the MMR provided by the functionalization would not last for beyond a few hours.

6.4.2 Carbohydrate-functionalized nanoparticles induced reduced antigen-specific CD8+ T cell expansion Thy1.1+ mice were subcutaneously immunized with a total of 100 μg of Ova, which for the particle groups consisted of 90 μg of soluble protein and 10 μg of protein encapsulated into

176

500 μg of the 20:80 CPTEG:CPH nanoparticles. One day prior to immunization all the mice were given 7.5 x 105 OT-I Thy1.2+ CD8+ T cells intravenously. This number of T cells was chosen based on preliminary studies, which showed that peripheral blood samples could be used to track OT-I cell proliferation using flow cytometry (data not shown). This adoptive transfer model enabled better evaluation of the antigen-specific CD8+ T cell response induced by the different nanoparticle formulations.

Peripheral blood was collected from each mouse at days 2, 6, and 10 PI to measure the expansion kinetics of the Ova-specific T cells in the circulation. Peripheral blood mononuclear cells (PBMCs) were labeled with anti-CD8ß and Thy1.2 (Figure 6.3A). The percentage of circulating endogenous CD8+ T cells within the lymphocyte gate varied from 10-15%. There were significantly fewer Ova-specific T cells, but they could easily be discerned using flow cytometry. The data indicated that the peak expansion of Ova-specific CD8+ T cells occurred at day 6 PI and by day 10 the contraction phase of the OT-I T cells had been completed (Figure

6.3B). In these studies, the non-functionalized nanoparticles and soluble Ova treatment groups induced the greatest expansion of CD8+ T cells in the peripheral blood.

The PBMCs were labeled with antibodies specific for the intracellular cytokines TNF-α and IFN-γ to characterize T cell activation (Figure 6.4A). Cells from immunized animals showed significantly greater amounts of pro-inflammatory cytokines relative to cells from animals that were administered saline (Figure 6.4B). As a percentage, the functionalized particles induced a similarly sized population of OT-I T cells that were double positive for TNF-α and IFN-γ as did treatment of the mice with the non-functionalized particles. However, there were fewer CD8+ T cells in the peripheral blood of the mice that received the functionalized particle treatments, and so the numbers of the TNF-α and IFN-γ producing cells are lower in these groups.

177

6.4.3 Carbohydrate-functionalized nanoparticles induced weak memory recall responses Four weeks PI, mice receiving OT-I Thy1.2+ T cells were challenged with EG7 ovalbumin expressing tumor cells. Four days following the challenge, peripheral blood was collected to analyze the ability of each vaccination group to mount a recall response. No significant differences were observed in the number of antigen-specific CD8+ T cells present in

100 μL of peripheral blood between any of the immunized animals. Both non-functionalized nanoparticles and soluble protein induced significantly more cells expressing TNF-α and IFN-γ

(Figure 6.5). In contrast, the carbohydrate-functionalized particles induced virtually no recall response as indicated by the percentage of OT-I T cells that were positive for the presence of

IFN-γ and TNF-α.

6.4.4 Nanovaccines induced robust antibody titers Thy1.1+ mice were immunized in the same manner and with the same treatment groups as previously described. However, one day prior to immunization all the mice were given 7.5 x

105 OT-II Thy1.2+ CD4+ T cells intravenously to assess the ability of the Ova-specific CD4+ T cells to enhance the antigen-specific IgG response. Peripheral blood was collected every two weeks and antibody titers were quantified using an ELISA (Figure 6.6A). The data indicated a dramatic increase in titer between 4 and 6 weeks. At each time point, the animals vaccinated with nanoparticles had higher titers than those in animals vaccinated with soluble Ova or animals treated with saline. In particular, it was observed that the animals vaccinated with the carbohydrate-functionalized nanoparticles showed titers that were similar to that induced in the animals that were vaccinated with the non-functionalized particles.

At day 42, the serum was analyzed for isotype switching. The vast majority of IgG secreted was IgG1 (Figure 6.6B & 6.6C), as expected. However, animals vaccinated with the non-functionalized nanoparticles, soluble protein, and di-mannose-functionalized nanoparticles

178

showed a small but significant production of IgG2c compared to mice receiving the saline treatment or linker-functionalized nanoparticles. Mice immunized with non-functionalized nanoparticles produced the most IgG2c.

6.5 Discussion Recent investigations have revealed that engaging CLRs on DCs can be a novel strategy to rationally design vaccines (27-30). Nanoparticles were designed to enhance APC activation by covalently linking carbohydrate ligands specific for the MMR onto their surface. The work described herein assessed the adaptive immune response in animals subcutaneously immunized with di-mannose-functionalized polyanhydride nanoparticles.

Both the non-functionalized nanoparticles and soluble Ova induced 10-fold greater expansion of CD8+ T cells, which was significantly higher than that induced by the carbohydrate-functionalized nanoparticles (Figure 6.3B). It is interesting to note that at day 2, more CD8+ T cells were recovered from the peripheral blood of animals treated with saline compared to that of any of the other vaccination groups. This suggests that at day 2, the T cells in the vaccinated groups receiving antigen are not in the peripheral blood, but more likely in the lymph node. By day 10, the OT-I T cells in the peripheral blood from all the animals studied

(independent of immunization) contracted to near baseline levels (Figure 6.3B). It is well known that the majority of effector T cells succumb to apoptosis when antigen levels fall (31). This observation is interesting in that even though the nanoparticles prolonged the release of antigen, the kinetics of the T cell response were not changed relative to that induced by administering soluble protein.

At peak expansion of an acute response, CD8+ T cells have a functional profile that includes production of large amounts of IFN-γ and TNF-α and low amounts of IL-2 (32). The

179

efficacy of the T cell responses was evaluated by analyzing the production of IFN-γ and TNF-α six days PI (Figure 6.4). Compared to animals that received saline, the T cells from immunized animals (soluble or particle-based) produced significantly higher amounts of IFN-γ and TNF-α.

In addition, the percentage of cytokines produced by the T cells of animals receiving the different immunization regimens was similar suggesting that the T cells were similarly activated in all the animals. However, given that the carbohydrate-functionalized particles induced less expansion of T cells than that induced by the non-functionalized particles or soluble protein, the total number of T cells expressing these cytokines was reduced. This ultimately suggests that the

CTL response generated in animals receiving these particles would be less efficacious than that induced in animals receiving the non-functionalized nanoparticles.

Most studies that have targeted CLRs use monoclonal antibodies directly conjugated to antigen, while others use carbohydrate-functionalized antigen (2; 33-36). In this work, carbohydrate-functionalized polyanhydride nanoparticles were administered to animals to analyze the adaptive immune response. Many trends observed in vivo by other investigators also translate to the carbohydrate-functionalized polyanhydride nanoparticles. For example, it has been observed that targeting antigen to DC receptors without maturation stimuli can result in tolerogenic or weaker CTL responses (36; 37). Successful targeting strategies for inducing CTLs combine targeting CLRs with other adjuvants, such as Poly(I:C) or anti-CD40 (38; 39). It is possible that using one of these adjuvants together with DC targeting would result in a synergistic T cell response.

In this regard, there are many examples in the literature of CLRs working synergistically with other adjuvants and TLR ligands to modulate the immune response. In one approach, dectin-1 stimulation by β-glucans with simultaneous administration of TLR2 or TLR4 agonists

180

increased the secretion of pro-inflammatory cytokines from primary monocytes and macrophages compared to the cytokine secretion when the ligands were administered individually (40). The importance of this synergy is further revealed by stimulating human DCs using nanoparticles co-encapsulating both TLR ligands and antigen and then coating those nanoparticles with DC-SIGN-specific antibodies. It was observed that co-encapsulation led to increased DC maturation and 100-fold enhancement of CTL responses compared to that induced by administering soluble TLR ligands alone (41). Given the diverse repertoire of PAMPs on pathogens, it is likely that an infection will stimulate multiple PRRs, making crosstalk and co- signaling essential to combating infection. It may be important to replicate this effect by using combination adjuvants in order to induce appropriate immune responses.

The ability of the nanovaccines to induce memory recall responses was tested by challenging the mice with Ova-expressing tumor cells 4 weeks PI (Figure 6.5). CD8+ T cell activation was investigated 4 days post challenge. The CD8+ T cells of animals treated with both the non-functionalized nanoparticles and soluble protein secreted significant amounts of IFN-γ and TNF-α, indicating the importance of a strong primary CD8+ T response to induce an efficacious memory recall response. In contrast, the weak primary CD8+ T cell responses in the animals immunized with the functionalized nanoparticles translated into no recall response.

The strength and type of the adaptive immune response generated uses cues from the innate immune response (42). APCs provide three essential signals to stimulate T cells, the first of which is presentation of the antigenic epitope on major histocompatibility complex (MHC) proteins (43). Signal 2 is the co-stimulatory molecules such CD80/86 or CD40 recognized by

CD28 or CD40L on the T cells, respectively. Signal 3 is provided by secreted cytokines generated by the APC. The T cell requires all three of these signals to become fully activated or

181

it may undergo apoptosis or become altered so that it cannot become activated (43). This is why live pathogens, such as viruses, are the most efficient primers of CTL responses. They have complex patterns that trigger several PRRs which induces cellular activation while simultaneously presenting antigen by stimulating all three of these signals (2).

Some studies have shown that the addition of IL-12 or Type I interferons enhance the proliferation and survival of activated T cells (44; 45). This has furthered the concept that signal

3, which is related to cytokine secretion, is very important for generating potent CD8+ T cell responses (46). Past in vitro investigations have shown that the carbohydrate-functionalized particles induced increased cytokine secretion that drove acute-phase responses, such as TNF-α,

IL-1β, and IL-6, but they did not induce significantly different IL-12 production compared with non-functionalized nanoparticles (7). This could potentially explain why the carbohydrate- functionalized particles generated a limited CD8+ T cell response.

Targeting antigen to CLRs on the surface of DCs is also intended to enhance CD4+ T cell responses and antibody-mediated responses. There is some discrepancy in the literature on whether targeting techniques require adjuvants to boost antibody titers (36). In targeting DEC-

205, Dectin-1, the MMR, or CLEC9A, very weak antibody responses were observed without the use of an adjuvant (47; 48). Upon use of Ova targeted to CLEC9A+ APCs, CD4+ T cells differentiated into FOXP3+ regulatory T cells without an adjuvant (47). Co-administration of an adjuvant altered the CD4+ response to a Th1 profile with Poly(I:C) or to a Th17 profile with curdlan. Other studies observed avid antibody titers by targeting CLRs without an adjuvant and the general consensus remains that strong antibody titers can be generated without an adjuvant

(49). In this study, the animals receiving nanovaccines showed robust antibody titers relative to animals receiving soluble protein (Figure 6.6A). At each time point, the animals immunized with

182

the carbohydrate-functionalized nanoparticles induced similar levels of antibody titers compared to that induced in animals immunized with non-functionalized nanoparticles. The surface- eroding polyanhydride nanoparticles provide a sustained release of antigen, ultimately enabling antigen persistence, which has been shown to be important for inducing long-lived plasma cells and B cells (10; 50). By creating an antigenic depot, all three types of nanoparticles induced the formation of particulate antigen, which is more easily recognized by B cells (9). Additionally, the antibody titer data reinforce the CD8+ T cell data. The predominant IgG subtype within the antibody production in animals receiving any type of vaccination in this study was observed to be IgG1 (Figure 6.6B and 6.6C). However, animals vaccinated with non-functionalized particles, soluble protein, and di-mannose-functionalized particles induced measurable amounts of IgG2c as well. The animals vaccinated with the linker-functionalized particles had the least IgG2c production, and correspondingly the least CD8+ T cell expansion. Both the T cell data and the antibody class switching observations suggest that carbohydrate functionalization did not polarize the T cell response to generate Th1 effector cells.

Figure 6.2 shows that the surface charge of the functionalized particles changes with time upon exposure to buffer. These results suggest that when the particles erode, they shed both ethylenediamine as well as the attached carbohydrate. As a result, labile surface modification may not provide a persistent “danger signal” that is needed to trigger PRRs, induce DC maturation, and activate T cells. The data in Figure 6.2 indicate that the particles are losing the targeting ligand within one day. It is known that DC maturation occurs in the first 24 hours, but in contrast, T cell activation requires multiple days (51). These data may also explain why the antibody responses were similar in the animals receiving the non-functionalized and

183

functionalized particles. B cell activation occurs in germinal centers, which begin to develop about 7-10 days after exposure to antigen (52).

Unexpectedly, the carbohydrate-functionalized nanoparticles did not upregulate the adaptive immune response. In vitro data from previous studies showed that the functionalized nanoparticles were internalized effectively by MØ and DCs (6-8). Additionally these cells showed signatures of enhanced activation upon incubation with carbohydrate functionalized nanoparticles (5). The discrepancy between the in vitro results from previous studies and the in vivo results from the current study can be attributed to several factors. As alluded to earlier, the sugar could erode from the surface of the nanoparticles before the DCs interact with the functionalized surface. Alternatively, the type of response from a cell can depend on the avidity with which the ligand binds to a receptor (53; 54). The in vivo environment behaves more like an infinite sink compared to wells in tissue culture plates (55). As a result, DCs in vivo may have reduced opportunities to interact with nanoparticles and their carbohydrate surfaces as they otherwise would in vitro. Many other cells express CLRs and scavenger receptors including those of the mononuclear phagocytic system (56; 57). This system is comprised of highly phagocytic macrophages whose job is to maintain homeostasis in tissues (57). As a result functionalized nanoparticles could be getting taken up by this system which means the antigen encapsulated within the nanoparticles may be degraded and not be presented to T cells. Say something about the MMR receptor and the need to present multiple signals here, as discussed at the vaccine meeting. Nonetheless, these studies generated important insights with respect to the in vivo immune response to carbohydrate functionalized nanoparticles. It is possible that inducing a more persistent danger signal, perhaps by conjugating polyanhydrides with carbohydrates prior to nanoparticle synthesis, will help enhance T cell activation in vivo.

184

6.6 Conclusions This study investigated the capabilities of carbohydrate-functionalized polyanhydride nanoparticles to induce adaptive immune responses in mice using adoptive transfer of antigen- specific T cells. Animals receiving non-functionalized nanoparticles and soluble protein formulations had more OT-I CD8+ T cell expansion compared to animals that received carbohydrate-functionalized nanoparticles. Additionally, mice receiving these particles showed a reduced ability to produce pro-inflammatory cytokines and had a weak memory recall response upon challenge with antigen-specific tumor cells. Nonetheless animals immunized with all the particle formulations, regardless of functionalization, showed robust antibody responses compared to animals that received soluble protein. It is hypothesized that the carbohydrate modification may not provide a persistent danger signal that is needed to trigger PRRs, induce

DC maturation, and activate T cells. It is possible that co-stimulation with another adjuvant or

TLR ligand could provide the maturation signals to synergize with the targeting of CLRs to induce robust T cell immunity.

6.7 Acknowledgements The authors are grateful for the assistance of Dr. Shawn Rigby of the Iowa State Flow

Cytometry Facility. Financial support is acknowledged from The US Army Medical Research and Material Command (Grant No. W81XWH-10-1-0806).

6.8 References 1. Fahmy TM, Demento SL, Caplan MJ, Mellman I, Saltzman WM. 2008. Design opportunities for actively targeted nanoparticle vaccines. Nanomedicine (Lond) 3:343-55

2. Kastenmuller W, Kastenmuller K, Kurts C, Seder RA. 2014. Dendritic cell-targeted vaccines- -hope or hype? Nature reviews. Immunology 14:705-11

3. Park CG. 2014. Vaccine strategies utilizing C-type lectin receptors on dendritic cells in vivo. Clinical and experimental vaccine research 3:149-54

185

4. Sancho D, Reis e Sousa C. 2012. Signaling by myeloid C-type lectin receptors in immunity and homeostasis. Annual review of immunology 30:491-529

5. Chavez-Santoscoy AV, Roychoudhury R, Pohl NL, Wannemuehler MJ, Narasimhan B, Ramer-Tait AE. 2012. Tailoring the immune response by targeting C-type lectin receptors on alveolar macrophages using "pathogen-like" amphiphilic polyanhydride nanoparticles. Biomaterials 33:4762-72

7. Goodman JT, Ramirez JEV, Boggiatto PM, Roychoudhury R, Pohl NLB, et al. 2014. Nanoparticle Chemistry and Functionalization Differentially Regulates Dendritic Cell- Nanoparticle Interactions and Triggers Dendritic Cell Maturation. Particle & Particle Systems Characterization 31:1269-80

8. Phanse Y, Carrillo-Conde BR, Ramer-Tait AE, Roychoudhury R, Pohl NL, et al. 2013. Functionalization of polyanhydride microparticles with di-mannose influences uptake by and intracellular fate within dendritic cells. Acta biomaterialia 9:8902-9

9. Haughney SL, Ross KA, Boggiatto PM, Wannemuehler MJ, Narasimhan B. 2014. Effect of nanovaccine chemistry on humoral immune response kinetics and maturation. Nanoscale 6:13770-8

11. Kipper MJ, Wilson JH, Wannemuehler MJ, Narasimhan B. 2006. Single dose vaccine based on biodegradable polyanhydride microspheres can modulate immune response mechanism. Journal of biomedical materials research. Part A 76:798-810

12. Vela Ramirez JE, Roychoudhury R, Habte HH, Cho MW, Pohl NL, Narasimhan B. 2014. Carbohydrate-functionalized nanovaccines preserve HIV-1 antigen stability and activate antigen presenting cells. Journal of biomaterials science. Polymer edition 25:1387-406

186

13. Carrillo-Conde B, Schiltz E, Yu J, Chris Minion F, Phillips GJ, et al. 2010. Encapsulation into amphiphilic polyanhydride microparticles stabilizes Yersinia pestis antigens. Acta biomaterialia 6:3110-9

14. Haughney SL, Petersen LK, Schoofs AD, Ramer-Tait AE, King JD, et al. 2013. Retention of structure, antigenicity, and biological function of pneumococcal surface protein A (PspA) released from polyanhydride nanoparticles. Acta biomaterialia 9:8262-71

15. Ulery BD, Kumar D, Ramer-Tait AE, Metzger DW, Wannemuehler MJ, Narasimhan B. 2011. Design of a protective single-dose intranasal nanoparticle-based vaccine platform for respiratory infectious diseases. PloS one 6:e17642

16. Ulery BD, Petersen LK, Phanse Y, Kong CS, Broderick SR, et al. 2011. Rational design of pathogen-mimicking amphiphilic materials as nanoadjuvants. Scientific reports 1:198

17. Torres MP, Vogel BM, Narasimhan B, Mallapragada SK. 2006. Synthesis and characterization of novel polyanhydrides with tailored erosion mechanisms. Journal of biomedical materials research. Part A 76:102-10

18. Conix A. 1966. Poly[1,3-bis(p-carboxyphenoxy)-propane anhydride]. Macromolecular Synthesis 2:95-8

19. Torres MP, Determan AS, Anderson GL, Mallapragada SK, Narasimhan B. 2007. Amphiphilic polyanhydrides for protein stabilization and release. Biomaterials 28:108-16

20. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

21. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

22. Jaipuri FA, Pohl NL. 2008. Toward solution-phase automated iterative synthesis: fluorous-tag assisted solution-phase synthesis of linear and branched mannose oligomers. Organic & biomolecular chemistry 6:2686-91

187

23. Song EH, Osanya AO, Petersen CA, Pohl NL. 2010. Synthesis of multivalent tuberculosis and Leishmania-associated capping carbohydrates reveals structure-dependent responses allowing immune evasion. Journal of the American Chemical Society 132:11428-30

24. Masuko T, Minami A, Iwasaki N, Majima T, Nishimura S, Lee YC. 2005. Carbohydrate analysis by a phenol-sulfuric acid method in microplate format. Analytical biochemistry 339:69-72

25. Huntimer L, Ross KA, Darling RJ, Winterwood N, Boggiatto PM, et al. 2014. Polyanhydride nanovaccine platform enhances antigen-specific cytotoxic T cell responses. Technology 2:171-5

26. Vela-Ramirez JE, Goodman JT, Boggiatto PM, Roychoudhury R, Pohl NL, et al. 2015. Safety and biocompatibility of carbohydrate-functionalized polyanhydride nanoparticles. The AAPS journal 17:256-67

27. Geijtenbeek TBH, Gringhuis SI. 2009. Signalling through C-type lectin receptors: shaping immune responses. Nature Reviews Immunology 9:465-79

28. Schreibelt G, Klinkenberg LJ, Cruz LJ, Tacken PJ, Tel J, et al. 2012. The C-type lectin receptor CLEC9A mediates antigen uptake and (cross-)presentation by human blood BDCA3+ myeloid dendritic cells. Blood 119:2284-92

29. Lu Y, Kawakami S, Yamashita F, Hashida M. 2007. Development of an antigen- presenting cell-targeted DNA vaccine against melanoma by mannosylated liposomes. Biomaterials 28:3255-62

30. Lam JS, Mansour MK, Specht CA, Levitz SM. 2005. A model vaccine exploiting fungal mannosylation to increase antigen immunogenicity. Journal of immunology 175:7496-503

31. Kaech SM, Wherry EJ, Ahmed R. 2002. Effector and memory T-cell differentiation: implications for vaccine development. Nature reviews. Immunology 2:251-62

32. Harty JT, Badovinac VP. 2008. Shaping and reshaping CD8+ T-cell memory. Nature reviews. Immunology 8:107-19

33. Bonifaz LC, Bonnyay DP, Charalambous A, Darguste DI, Fujii S, et al. 2004. In vivo targeting of antigens to maturing dendritic cells via the DEC-205 receptor improves T cell vaccination. The Journal of experimental medicine 199:815-24

188

34. Wijagkanalan W, Kawakami S, Takenaga M, Igarashi R, Yamashita F, Hashida M. 2008. Efficient targeting to alveolar macrophages by intratracheal administration of mannosylated liposomes in rats. Journal of controlled release : official journal of the Controlled Release Society 125:121-30

35. Tsuji T, Matsuzaki J, Kelly MP, Ramakrishna V, Vitale L, et al. 2011. Antibody-targeted NY-ESO-1 to mannose receptor or DEC-205 in vitro elicits dual human CD8+ and CD4+ T cell responses with broad antigen specificity. Journal of immunology 186:1218-27

36. Tacken PJ, Figdor CG. 2011. Targeted antigen delivery and activation of dendritic cells in vivo: steps towards cost effective vaccines. Seminars in immunology 23:12-20

37. Lindquist RL, Shakhar G, Dudziak D, Wardemann H, Eisenreich T, et al. 2004. Visualizing dendritic cell networks in vivo. Nature immunology 5:1243-50

38. Longhi MP, Trumpfheller C, Idoyaga J, Caskey M, Matos I, et al. 2009. Dendritic cells require a systemic type I interferon response to mature and induce CD4+ Th1 immunity with poly IC as adjuvant. The Journal of experimental medicine 206:1589-602

39. Hawiger D, Inaba K, Dorsett Y, Guo M, Mahnke K, et al. 2001. Dendritic cells induce peripheral T cell unresponsiveness under steady state conditions in vivo. The Journal of experimental medicine 194:769-79

40. Ferwerda G, Meyer-Wentrup F, Kullberg BJ, Netea MG, Adema GJ. 2008. Dectin-1 synergizes with TLR2 and TLR4 for cytokine production in human primary monocytes and macrophages. Cellular microbiology 10:2058-66

41. Tacken PJ, Zeelenberg IS, Cruz LJ, van Hout-Kuijer MA, van de Glind G, et al. 2011. Targeted delivery of TLR ligands to human and mouse dendritic cells strongly enhances adjuvanticity. Blood 118:6836-44

42. Guermonprez P, Valladeau J, Zitvogel L, Thery C, Amigorena S. 2002. Antigen presentation and T cell stimulation by dendritic cells. Annual review of immunology 20:621- 67

43. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. 2002. Molecular Biology of the Cell. New York, NY: Garland Science, Taylor & Francis Group. 1463 pp.

189

44. Curtsinger JM, Lins DC, Mescher MF. 2003. Signal 3 determines tolerance versus full activation of naive CD8 T cells: dissociating proliferation and development of effector function. The Journal of experimental medicine 197:1141-51

45. Curtsinger JM, Valenzuela JO, Agarwal P, Lins D, Mescher MF. 2005. Type I IFNs provide a third signal to CD8 T cells to stimulate clonal expansion and differentiation. Journal of immunology 174:4465-9

46. Valenzuela J, Schmidt C, Mescher M. 2002. The roles of IL-12 in providing a third signal for clonal expansion of naive CD8 T cells. Journal of immunology 169:6842-9

47. Joffre OP, Sancho D, Zelenay S, Keller AM, Reis e Sousa C. 2010. Efficient and versatile manipulation of the peripheral CD4+ T-cell compartment by antigen targeting to DNGR-1/CLEC9A. European journal of immunology 40:1255-65

48. Boscardin SB, Hafalla JC, Masilamani RF, Kamphorst AO, Zebroski HA, et al. 2006. Antigen targeting to dendritic cells elicits long-lived T cell help for antibody responses. The Journal of experimental medicine 203:599-606

49. He LZ, Crocker A, Lee J, Mendoza-Ramirez J, Wang XT, et al. 2007. Antigenic targeting of the human mannose receptor induces tumor immunity. Journal of immunology 178:6259- 67

50. Elgueta R, de Vries VC, Noelle RJ. 2010. The immortality of humoral immunity. Immunological reviews 236:139-50

51. Murphy K. 2012. Janeway's Immunobiology. New York, NY: Garland Science, Taylor & Francis Group. 868 pp.

52. Victora GD, Nussenzweig MC. 2012. Germinal Centers. Annual Review of Immunology, Vol 30 30:429-57

53. Anderton SM, Radu CG, Lowrey PA, Ward ES, Wraith DC. 2001. Negative selection during the peripheral immune response to antigen. The Journal of experimental medicine 193:1-11

54. Zinkernagel RM, Hengartner H. 2001. Regulation of the immune response by antigen. Science 293:251-3

190

55. Kinnunen HM, Sharma V, Contreras-Rojas LR, Yu Y, Alleman C, et al. 2015. A novel in vitro method to model the fate of subcutaneously administered biopharmaceuticals and associated formulation components. Journal of controlled release : official journal of the Controlled Release Society 214:94-102

56. Murray PJ, Wynn TA. 2011. Protective and pathogenic functions of macrophage subsets. Nature reviews. Immunology 11:723-37

57. Chow A, Brown BD, Merad M. 2011. Studying the mononuclear phagocyte system in the molecular age. Nature reviews. Immunology 11:788-98

6.9 Figures

Figure 6.1 Schematic of polyanhydride chemistry, surface functionalization, and in vivo procedure. A) Structure of monomer diacids 1,6-bis(p-carboxyphenoxy)hexane (left) and 1,8- bis(p-carboxyphenoxy)-3,6-dioxaoctane (right). B) The nanoparticle surface was functionalized with a glycolic acid cap (middle) or with di-mannose (right) using an ethylenediamine linker.

191

Figure 6.2 Zeta potential kinetics of non-functionalized and linker-functionalized nanoparticles.

The ethylenediamine linker results in positively charged nanoparticles. After 4 h of incubation in

PBS at 37°C, the zeta potential of the particles becomes negative and by 24 h the zeta potential of the particles reach a plateau. In contrast, the zeta potential of the non-functionalized nanoparticles does not change with time.

192

Figure 6.3 Expansion and contraction of CD8+ OT-I T cells. A) Representative flow cytometry gating strategy. Lymphocytes were gated first and then the transferred T cell population was identified by gating for Thy1.2 and CD8b. * indicates significance from the saline population (p

< 0.05). B) Peripheral bleeds revealed that OT-I T cells for each treatment group expanded from day 2 to day 6 and contracted from day 6 to day 10 with soluble and non-functionalized particle treatment groups inducing the most T cell proliferation.

193

Figure 6.4 Intracellular cytokine production by CD8+ OT-I T cells at day 6. Intracellular cytokine staining was performed on OT-I T cells at their peak populations, which was day 6. A)

OT-I T cells from immunized groups strongly expressed IFN-γ. All treatment groups were significant from the saline populations with the most cytokine production being induced in the animal receiving the soluble antigen and the non-functionalized particle formulations. B) OT-I T cells from immunized groups readily produced TNF-α in addition to IFN-γ. C) Representative flow cytometry contour plots for each treatment group. * indicates significance from the saline population (p < 0.05).

194

Figure 6.5 Memory recall response of CD8+ OT-I T cells 4 days after tumor challenge. Four weeks post immunization, mice were challenged with Ova-expressing tumor cells. Peripheral bleeds four days after challenge were performed to see which formulations induced T cell memory. A) The number of OT-I cells in each peripheral bleed for each treatment group was similar. B) & C) Mice from both soluble and non-functionalized groups produced significantly more IFN-γ and TNF-α compared to that produced by the saline-treated mice. * indicates significance from the saline population (p < 0.05).

195

Figure 6.6 Antibody response at select time points following immunization and adoptive transfer of OT-II T cells. Mice that received OT-II T cells that were then immunized with one of the treatment groups were bled at days 14, 28, and 42 to analyze antibody titers. At each time point, immunized mice produced significantly more total IgG than saline mice. The particle formulations also outperformed the soluble protein. B) and C) Little class switching was induced with the day 42 titers as most of the antibody produced was IgG1. Animals receiving the soluble, non-functionalized, and di-mannose-functionalized nanoparticle immunization produced significantly more IgG2c compared to that of animals receiving saline. * indicates significance from the saline population (p < 0.05). # indicates significance from the soluble population (p <

0.05).

196

6.10 Tables Table 6.1 Nanoparticle size and characterization. Average Surface composition by EDS Sugar Group Diameter Average % Average % Average % Density (nm) C N O (μg/mg) Non-functionalized 177 ± 39 79.5 ± 2.9 0.04 ± 0.1 20.5 ± 2.8 Linker 186 ± 33 73.7 ± 2.7 5.54 ± 1.5 20.8 ± 1.5 Di-mannose 179 ± 44 71.6 ± 2.5 7.00 ± 1.7 21.4 ± 1.1 13.4 ± 2.59

Nanoparticles were sized with ImageJ of SEM files using 200 particles to record the average diameter. Nanoparticle functionalization was characterized by EDS and a phenol-sulfuric acid assay. EDS was used to identify an increase in nitrogen content to confirm surface attachment of the sugar and ethylenediamine. This is presented as the mean ± SD through seven independent readings. Mean ± SD of the sugar density was determined by a phenol-sulfuric acid assay measured in triplicate.

197

CHAPTER 7: HIGH THROUGHPUT AUTOMATED SYNTHESIS OF

PROTEIN-LOADED POLYANHYDRIDE NANOPARTICLE LIBRARIES

Jonathan T. Goodman1, Lucas Dunshee1, Akash Mitra1, and Balaji Narasimhan1

To be Submitted to ACS Combinatorial Sciences December, 2015

1 Department of Chemical & Biological Engineering, Iowa State University, Ames, Iowa 50011

198

7.1 Abstract A valuable resource in rapidly synthesizing and screening nanoscale biomaterial-based carriers for drug and vaccine delivery involves the development of high throughput techniques to make combinatorial libraries. This study describes the development of an automated high throughput robot for synthesizing polyanhydride nanoparticles encapsulating protein antigens.

Polyanhydrides are a class of safe and biodegradable polymers that have been widely used as drug and vaccine delivery vehicles. The robot contains a multiplexed homogenizer and has the capacity to handle parallel streams of monomer or polymer solutions to synthesize both polymers and/or nanoparticles. Copolymer libraries were synthesized using the momomers, sebacic acid

(SA), 1,6-bis(p-carboxyphenoxy)hexane (CPH), and 1,8-bis(p-carboxyphenoxy)-3,6-dioxactane

(CPTEG). Nanoparticle libraries of varying copolymer compositions encapsulating the model antigen, ovalbumin, were synthesized using flash nanoprecipitation. Concentrations of the surfactant, Span 80, were varied to test its effect on protein encapsulation efficiency as well as protein release kinetics. It was observed that the surfactant did not significantly affect protein release rate, but its presence enhanced protein encapsulation efficiency. Particles synthesized with the high throughput method were compared to particles made with conventional anti- solvent nanoprecipitation in order to validate the high throughput approach. Protein burst and release kinetics between the two methods were similar, even though particles synthesized using the high throughput technique were observed to be smaller. Finally, it was demonstrated that the high throughput method could be adapted to functionalize the surface of the particles to aid in the design of targeted drug and vaccine delivery systems. These results suggest that the newly developed automated high throughput system is a viable alternative to conventional methods for studying and screening vaccine and drug delivery vehicles.

199

7.2 Introduction The concept of combinatorial libraries and high throughput screening gained widespread support in the 1980s when the pharmaceutical industry successfully pioneered these techniques to synthesize and screen small molecules that would fill their drug pipelines (1). High throughput technologies are needed because of limitations in our ability to make accurate predictions about the activity and efficacy of drugs or drug delivery vehicles. They allow for a greater number of compounds to be screened enabling the empirical selection of a drug or drug delivery vehicle with the most favorable properties. Combinatorial libraries are advantageous relative to conventional methods because they have the potential to integrate multiple components including the synthesis and the screening process together reducing the overall workload (2; 3).

The key challenge remains in the actual design step to ensure that high throughput mechanisms and automation equipment most efficiently screen and select drugs or drug delivery vehicles for their optimal properties.

Combinatorial methods have been used to optimize biomaterials design and drug delivery vehicles for many complex biomedical end-use applications, including targeting cancer, optimizing gene delivery, penetrating the blood brain barrier (BBB), and modulating the immune response (2-5). In particular, techniques that automate and accelerate the synthesis of libraries of polymeric nanoparticles and microparticles for use as drug delivery vehicles are of great interest.

This is because there are many complex interactions between the drug, the delivery vehicle, and the cells or tissue interacting with the delivery vehicle (6; 7). High throughput technologies have the ability to rapidly screen these interactions to identify and select those that are favorable.

Previous work from our laboratories and other research groups has utilized these technologies to study the effect of polymer composition and polymer geometry on protein release (7-13). The

200

activation of macrophages and the biocompatibility of cells with polyanhydride libraries were successfully evaluated (14). These studies demonstrated that polyanhydride formulations have excellent properties as drug delivery vehicles owing to their ability to stabilize proteins, to provide sustained release of a protein or drug (driven by the surface erosion mechanism of polyanhydrides), and to target specific cells (15-18). Additionally, polyanhydride chemistry is highly tunable (i.e., by varying copolymer composition), which allows the rate of drug release to be modulated. Finally, polyanhydride nanoparticle libraries efficiently activated macrophages without inducing any toxicity (14; 19; 20).

This work builds upon the previous studies via the design of a new and automated high throughput system for synthesizing and characterizing both polymer libraries as well as protein- loaded nanoparticle libraries. The high throughput robot described herein efficiently synthesized polymer libraries, synthesized and functionalized nanoparticle libraries, and enabled rapid analysis of protein release kinetics from these libraries. The robot was also used to study the effect of several other parameters on nanoparticle size and protein release, including sonication conditions, use of surfactants, etc. All the high throughout data was validated using conventional syntheses and protein release studies.

7.2 Materials and Methods

7.2.1 Materials Sebacic acid monomer was purchased from Sigma Aldrich (St. Louis, MO). 4-p- hydroxybenzoic acid, 1-methyl-2-pyrrolidinone, 1,6-dibromohexane, and triethylene glycol were also purchased from Sigma Aldrich and they were used for CPTEG and CPH monomer synthesis. Monomer and polymer synthesis utilized dimethyl formamide, sulfuric acid, acetonitrile, toluene, potassium carbonate, and acetic anhydride, all of which were obtained from

201

Fisher Scientific (Fairlawn, NJ). Apollo Scientific (Cheshire, UK) was the source for 4-p- fluorobenzonitrile used in monomer synthesis. Nanoparticle synthesis required pentane and methylene chloride from Fisher Scientific as well as the surfactant Span 80 from Sigma Aldrich.

Deuterated chloroform from Cambridge Isotope Laboratories (Andover, MA) was purchased for

1H NMR analysis. Ovalbumin (Ova) was purchased from Sigma-Aldrich passed through a

Thermo Scientific Detoxi-Gel Endotoxin removing column before use to ensure that the endotoxin was removed. 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N- hydroxysuccinimide (NHS) from Fisher Scientific, ethylenediamine from Sigma Aldrich, and glycolic acid from Acros Organics (Geel, Belgium) were used for nanoparticle functionalization.

7.2.2 Automated high throughput robot In these studies, a custom built robot called the Nanoprep 24 (OMNI International,

Kennesaw, GA) was used that allows for automated and facile preparation of polymers and nanoparticle libraries (Figure 7.1). The robot has three syringe pumps, permitting individual stock solutions of monomers or polymers to be pumped without contaminating each other

(Figure 7.1C). Two of the syringe pumps lead to stationary dispensing nozzles, and the third syringe pump is connected to a stainless steel pipette nozzle to either handle liquid dispensing or for programmed liquid transfers. The robot is equipped with a 20 kHz ultrasonic probe that can range in amplitude from 0-100%. There are 24 sampling locations in each of its two racks. The first rack is able to hold glass test tubes up to 15 mm in diameter and 10 mL in volume (Figure

7.1D). It is removable and is capable of withstanding the higher temperatures typically used in polymer synthesis (Figure 7.1E). The second rack is fitted to hold 50 mL conical polypropylene tubes. It is also removable and is designed to be placed in a -80°C freezer. It can accommodate a dry ice payload to maintain cool temperatures during nanoparticle synthesis. Finally, there is a waste and rinse rack, where the robot can dispose of supernatants and rinse itself to prevent

202

contamination with other samples in the rack. The robot is connected to a computer through a

USB port which serves as the main user interface. Spreadsheets allow for inputting process variables, such as liquid volumes, into each sample rack or sonication duration and amplitude of each sample.

7.2.3 Polymer library synthesis 1,6-bis(p-carboxyphenoxy)hexane (CPH) and 1,8-bis(p-carboxyphenoxy)-3,6- dioxaoctane (CPTEG) diacids were synthesized as described previously (16; 21-23). In conventional synthesis protocols, these diacids were added to a round bottom flask at appropriate molar ratios of CPTEG and CPH. Excess acetic anhydride was added and the flask was submerged in an oil bath at 125°C for 30 min to acetylate the monomers. The acetic anhydride was removed by rotary evaporation and the round bottom flask was re-submerged into an oil bath at 140°C to begin polymerization. A vacuum (<0.3 torr) was applied for five hours to carry out the polymerization. At the end of polymerization, the round bottom flask was brought to standard temperature and pressure, and methylene chloride was added to dissolve the polymer.

The polymer was precipitated by adding the methylene chloride solution dropwise to excess hexane. This hexane solution was kept chilled by a dry ice bath. The polymer was allowed to settle to the bottom of the hexane solution allowing the hexane to be poured off. The remaining polymer was dried under vacuum. The CPH and SA monomers were acetylated separately as described previously (23; 24). Conventional synthesis of CPH:SA copolymers was nearly identical to that of CPTEG:CPH copolymer synthesis except the oil bath was at 180°C and the reaction proceeded for 30 min. Precipitation was carried out at room temperature in hexane.

High-throughput synthesis of CPTEG:CPH and CPH:SA libraries was carried out simultaneously by creating 15 mg/mL stock solutions of CPH diacid, CPTEG diacid, and SA diacid in acetic anhydride. The diacids were acetylated by incubating in a 125°C mineral oil bath

203

for 30 min. Each solution was connected to its own 1/8” diameter tubing on the robot and the robot was programmed to individually add specific amounts of acetylated monomer to 10 mL glass test tubes (VWR International, Radnor, PA). The test tubes were removed from the robot and placed in a 180°C vacuum for 3 h for CPH:SA copolymers and 5 h for CPTEG:CPH copolymers. The acetic anhydride gradually evaporated into a chilled bump trap and the acetylated monomers were allowed to copolymerize.

Polymer structure, molecular weight, and purity were verified using 1H NMR spectroscopy on a Varian VXR 300 MHz spectrometer. Differential scanning calorimetry (DSC,

Q2000, TA Instruments, New Castle, DE) was used to determine thermal properties (i.e., glass transition temperature and melting point).

7.2.4 Nanoparticle library synthesis Conventional synthesis of protein-loaded nanoparticles was carried out as described previously (19). Briefly, polymer and 1% (w/w) Ova were weighed into a scintillation vial and dissolved in methylene chloride at a concentration of 25 mg/mL. A surfactant, Span 80, was added to the polymer solution in varying amounts depending on the copolymer chemistry. The polymer solution was then sonicated at 40% amplitude for 30 s and precipitated into a non- solvent, pentane, at an anti-solvent:solvent ratio of 80:1. The synthesis of CPH:SA nanoparticles was carried out at room temperature, while that of CPTEG:CPH nanoparticles was carried out in a 4°C cold room with a pentane bath chilled to -20°C. Nanoparticles were collected through vacuum filtration.

For high throughput nanoparticle library synthesis, methylene chloride containing 1%

(w/w) Ova and the correct amount of Span 80 was added via one of the syringe pumps to each polymer containing test tube. The robot sonicated each test tube at 40% amplitude for 30 s to

204

facilitate polymer dissolution and disperse the protein. The polymer solution in each test tube was then transferred to its corresponding 50 mL conical tube filled with 45 mL of pentane using the stainless steel pipette nozzle. The polymer concentration in the methylene chloride solution was 25 mg/mL and the nanoprecipitation maintained a solvent: anti-solvent ratio of 1:80. Since the formation of CPTEG:CPH nanoparticles required low temperatures, the pentane was previously chilled in the -80°C freezer. In addition, dry ice and liquid nitrogen were added to the holding rack to help keep the polymer at -20°C during nanoparticle synthesis. Conical tubes of identical treatment groups were combined and particles were collected by vacuum filtration to acquire enough particles for release studies. The nanoparticles were characterized by scanning electron microscopy (SEM, FEI Quanta 250, Hillsboro, OR) and an average size distribution was generated using ImageJ image analysis software from the National Institutes of Health

(Bethesda, MD) with 200 nanoparticles per image.

7.2.5 Nanoparticle library functionalization The particles were functionalized with a modified lactose, a modified galactose, or glycolic acid using an ethylenediamine linker. The modified sugars contained a carboxylic acid group to link the sugar to the ethylenediamine through an amide bond (25). Mercury(II)- catalyzed allylation of penta-O-acetyl-lactosyl bromide was used to prepare β-1-O-allylated lactose (26; 27). This was oxidized with excess NaIO4 in the presence of RuCl3 to generate the acid (28). Deprotection of the acetyl protecting groups occurred with K2CO3 to yield the final product. Similarly, β-1-O-allylated galactose was derived from β-penta-O-acetylated galactose in the presence of allyl alcohol and BF3OEt2 (18; 26). The carboxylic acid was generated from oxidation of the allyl handle with NaIO4 and RuCl3 and the acetyl deprotecting groups were removed using K2CO3.

205

The robot was used to perform surface functionalization of the nanoparticle libraries. For these studies, 10 mg of nanoparticles were added to 1.5 mL centrifuge tubes and loaded into the test tube holder of the robot. A 45 mg/mL stock solution of EDC and a 62 mg/mL stock solution of NHS was prepared and connected to the robot. Functionalization was carried out using a two- step reaction. The first reaction involved pumping 10 mole equivalents of EDC and 12 mole equivalents of NHS into each test tube of the nanoparticle library. The robot sonicated each suspension for 30 s at 30% amplitude. As the robot progressed through the sonication step, 10 mole equivalents of ethylenediamine were added manually. The library was then removed and placed on a tube rotor at 4°C for 1 h. Centrifugation for 10 min at 10,000 g pelleted the nanoparticles and the test tubes were placed back on the robot to extract the supernatant. The robot then added fresh nanopure water to wash the particles and sonicated the suspension for 30 s at 30% amplitude. The nanoparticle library was centrifuged for another 10 min at 10,000 g. The robot performed the second step of the reaction by extracting the supernatant followed by adding

10 mole equivalents of EDC and 12 mole equivalents of NHS into each test tube of the library.

Sonication for 30 s at 30% amplitude then continued, followed by manual addition of 10 equivalents of the sugar. The library was then removed from the robot and placed on the tube rotor for 1 h at 4°C. The nanoparticles were washed using the robot as previously described before being placed in a vacuum oven to dry for an additional one hour.

7.2.6 Characterization of functionalized nanoparticle libraries A phenol-sulfuric acid assay was performed to quantify the sugar density on the surface of the nanoparticle library using flat bottom 96-well plates (29; 30). The assay proceeded for 30 min at 95 °C at which time the reaction induced a color change in the solution. The absorbance of each well was measured at a wavelength of 490 nm using a SpectraMax M3 spectrophotometer (Molecular Devices, Sunnyvale, CA). This was compared against soluble

206

galactose and lactose standards, while non-functionalized nanoparticles were used as a negative control. Energy dispersive X-ray spectroscopy (EDS) was performed to assay for nitrogen content from the ethylenediamine.

To track the zeta potential of the functionalized particles as a function of time, five mg of the functionalized nanoparticles and 1 mL of PBS were placed in a 1.5 mL centrifuge tube, sonicated, and incubated at 37°C in a 100 rpm rotator. 40 μL of particle suspension was removed at each time point and diluted in 960 μL of cold nanopure water for a final concentration of 200

μg/mL. The zeta potential was then measured using a ZetaSizer Nano (Malvern, Worchester,

UK).

7.2.7 Protein release kinetics Eight mg of Ova-loaded nanoparticles were placed in 1.5 mL microcentrifuge tubes and suspended in 1 mL of 0.1 M PBS solution (pH = 7.4). Sonication was used to disperse the particles in the solution, after which the particles were placed in a 37°C incubator agitating at

100 rpm. At selected time points the nanoparticles were pelleted by centrifugation and the supernatants were extracted for protein quantification using a microBCA assay (ThermoFisher).

Fresh PBS was added to replace the removed supernatant maintaining a perfect sink. The Ova release kinetics are presented as a cumulative mass fraction as a function of time, which was determined by normalizing protein release for each time point with the total protein encapsulated, which was determined using a base extraction as described previously (31).

7.3 Results & Discussion

7.3.1 Polymer and nanoparticle characterization This work described the design and validation of a novel high throughput automated robot for efficient synthesis of libraries of polymers and protein-loaded nanoparticles for rapid

207

screening for properties that optimize their use as drug delivery vehicles. The robot enabled synthesis of polymer libraries whose properties were consistent with that of conventionally synthesized polymers, but with much more ease and with significant time savings. For example, in conventional synthesis, the reacted polymer is dissolved in methylene chloride so that it can be removed from the round bottom flask and is precipitated in hexanes. The polymer is filtered, dried and stored until it is used for nanoparticle synthesis. In the high throughput method, the number of steps required to convert monomer or prepolymer to nanoparticles are drastically reduced and produce significant time savings. Herein, the monomer or prepolymer and its solvent are pumped directly to its corresponding test tube. This test tube is then baked under vacuum where the solvent evaporates away into a chilled dewar collection flask to enable polymerization to proceed. This test tube is then stored until it is used for nanoparticle synthesis avoiding the need for a precipitation step in hexanes. Moreover, the high throughput protocol does not utilize the rotary evaporator during conventional synthesis resulting in further time savings. It is estimated that synthesizing a single CPTEG:CPH copolymer using conventional procedures requires approximately 2-3 h of laboratory time with an additional 18-20 h required for the monomers to polymerize, for the copolymer to dissolve in methylene chloride, and for the precipitated copolymer to dry. In contrast, the high throughput robot can synthesize an entire library of copolymers ranging in composition within 1-2 h of laboratory time and an additional 5 h for the monomers to polymerize with no time spent on precipitating the polymer. As a result, the high throughput synthesis process is more efficient relative to the conventional procedure as more copolymers compositions are synthesized simultaneously.

Representative copolymers were synthesized with the high throughput method and were

1 characterized using H NMR and DSC (Table 7.1). It was observed that Tg gradually increased

208

as the CPH content within the CPTEG:CPH copolymer increased. No defined melting point was observed for any of these copolymers, in agreement with previous studies that showed that these copolymers were amorphous (21). Additionally, increasing the CPH composition generally increased the molecular weight of the CPTEG:CPH copolymers. This suggests that the CPH monomer is more reactive than the CPTEG monomer. The semi-crystalline CPH:SA copolymers showed well-defined Tg’s and Tm’s. Both the Tg and Tm increased as the SA content increased within the copolymer.

In order to validate the high throughput method, nanoparticles were synthesized using both methods for only the 20:80 CPH:SA and 20:80 CPTEG:CPH copolymers. Different amounts of the surfactant, Span80, were incorporated into the synthesis procedures in order to investigate the effect of a surfactant on nanoparticle morphology. Both methods utilized a solid/oil/oil double emulsion technique to synthesize the nanoparticles. In both synthesis methods, as the amount of Span 80 was increased from 0% to 1%, the 20:80 CPH:SA nanoparticles became smaller and more monodisperse (Table 7.2, Figure 7.2). In contrast, the average diameter of the 20:80 CPTEG:CPH nanoparticles did not significantly change with the amount of Span 80 used. This is because increasing the amount of surfactant tended to induce aggregation in these nanoparticles. The CPH:SA polymers generally have much higher Tg’s (21;

32), resulting in precipitated nanoparticles that are not aggregated allowing it to withstand higher concentrations of surfactant. The 20:80 CPTEG:CPH nanoparticles synthesized using both methods had comparable sizes. However, the 20:80 CPH:SA nanoparticles synthesized by conventional methods were slightly larger than equivalent nanoparticles synthesized by the high throughput method (Table 7.2). The high-throughput 20:80 CPH:SA nanoparticles were smaller than the conventional nanoparticles, because the nozzle on the robot helps reduce solvent droplet

209

size preventing polymer from aggregating as it is exposed to the anti-solvent. However, particle size can be controlled by adjusting the nozzle velocity as shown by Figure 7.4.

Additionally, when using the high throughput method to synthesize the nanoparticles, it was observed that the nanoparticle size was affected by the transfer rate of the polymer solution to the anti-solvent during nanoprecipitation. As the transfer rate was increased, the nanoparticles became smaller. Operating the pump at its maximum nozzle velocity of 1200 μL/s resulted in nanoparticles that were on average 80 nm smaller than nanoparticles synthesized by much lower pump speeds (Figure 7.3). These observations indicate that the pump rate can be easily varied in the high throughput synthesis method, providing another tunable parameter to optimize nanoparticle size.

7.3.2 Protein release from polyanhydride nanoparticles The release kinetics of protein antigen from a delivery vehicle can greatly influence the immune response (33-36). In these studies, 1% (w/w) Ova was encapsulated into 20:80

CPTEG:CPH and 20:80 CPH:SA nanoparticles and the encapsulation efficiency and release kinetics were investigated as a function of surfactant concentration. Nanoparticles of the same compositions were compared to those synthesized using conventional methods in order to validate the high throughput process. Consistent with previous studies, polymer chemistry dictated the protein release rate from the nanoparticle library (37; 38). The hydrophobic 20:80

CPTEG:CPH nanoparticles displayed the slowest protein release rate, taking over 80 days to completely release the protein (ascertained by not detecting any protein in the supernatant using the microBCA assay). At early time points, the 20:80 CPTEG:CPH nanoparticles also had the smallest protein burst with only about 20-25% of the protein being released initially. These observations were consistent when using particles synthesized by both conventional and high

210

throughput methods. In contrast, the less hydrophobic 20:80 CPH:SA nanoparticles had significantly faster release rates and >90% of the encapsulated protein was released in the first several days from the particles synthesized by both methods. This formulation also had the largest protein burst as 70-80% of the protein was released in the first few time points. With its backbone consisting of aromatic rings and hexane, CPH is more hydrophobic relative to SA or

CPTEG. As a result polymer chemistries rich in CPH are more hydrophobic and have the slowest protein release rates (15). Polymer chemistry also significantly affected protein encapsulation efficiency (Table 7.3). Regardless of surfactant incorporation, the 20:80 CPTEG:CPH nanoparticles consistently had higher encapsulation efficiencies when compared to that of the

20:80 CPH:SA nanoparticles. This is likely due to the presence of the less hydrophobic CPTEG monomer, which contains hydrophilic ethylene glycol moieties, and has been shown to stabilize proteins in previous studies (16).

Incorporating the surfactant into the nanoparticles had a weak effect on protein release.

The protein burst was slightly increased when 0.2% Span 80 was used to synthesize 20:80

CPTEG:CPH nanoparticles, but the slope of the release rate was nearly identical compared to

20:80 CPTEG:CPH nanoparticles synthesized in the absence of any surfactant (Figure 7.4). The protein burst and the protein release rate was nearly unchanged when incorporating 0.2% Span

80 or 1% Span 80 into 20:80 CPH:SA nanoparticles. Incorporating surfactant into the nanoparticles did affect the protein encapsulation efficiency. As expected, adding surfactant increased the amount of protein encapsulated within the nanoparticles for both polymer chemistries (Table 7.3). This is because the surfactant enhances the stability of the insoluble protein as it is dispersed in the methylene chloride oil phase just prior to nanoprecipitation. Span based surfactants have been extensively used for stabilization of emulsions (39). Surfactant

211

function is quantified by its hydrophilic-lipophilic balance (HLB) value. Generally, surfactants with an HLB <10 tend to be soluble in oil phases, and an HLB between 3-6 is optimal for oil emulsifying agents (40). With an HLB of 4.3, Span 80 has been shown to serve as a useful emulsifying agent in oil emulsions (41).

The protein release rate and initial protein burst were comparable for both the high throughput and conventional synthesis methods. This validates the high throughput method in that it can reproducibly synthesize protein-loaded nanoparticles that have the same release kinetics as the conventional method. This agreement between the two methods also indicates that the small decrease in nanoparticle size synthesized using the high throughput method did not substantially affect protein release rate. The encapsulation efficiencies for both synthesis techniques exhibited similar trends in that incorporating Span 80 into the nanoparticles increased protein encapsulation efficiency. And finally, the CPTEG-rich formulation had the greatest encapsulation efficiencies when using both synthesis techniques.

7.3.3 Nanoparticle surface functionalization and characterization The robot was demonstrated to successfully conjugate modified sugars to the carboxylic acid end groups on the surface of the nanoparticle library using the same two-step reaction as in the conventional functionalization method. The chemical conjugation step was automated with the robot allowing for facile construction of surface-functionalized nanoparticle libraries to examine the effect of different sugars. The surface of 20:80 CPH:SA nanoparticles synthesized using the robot was functionalized with different sugars using the high throughput methodology

(Table 4.4). Nanoparticle libraries were chemically conjugated with glycolic acid (linker) and two different sugars (lactose and galactose) and characterized so that comparisons could be made to functionalized particles synthesized conventionally. The phenol-sulfuric acid assay revealed

212

that the amounts of lactose and galactose conjugated to 20:80 CPH:SA nanoparticles were comparable to the amounts of sugar on these particles when synthesized using conventional methods (18; 27; 30; 42). Energy dispersive spectroscopy (EDS) characterization also confirmed successful conjugation, as evidenced by the statistically significant nitrogen composition relative to that of the non-functionalized nanoparticles due to the ethylenediamine linker that couples the sugar to the nanoparticle.

In order to characterize the duration of time that the linker stays on the particle in solution, glycolic acid functionalized 20:80 CPH:SA and 20:80 CPTEG:CPH nanoparticles were synthesized using both methods. The time evolution of their surface potentials upon incubation in PBS was compared (Figure 7.5). At initial time points the functionalized nanoparticles of both chemistries had a positive zeta potential of ~+14 mV. Incubation in aqueous media led to a breakdown of the linkage to the surface of the nanoparticles and the zeta potential of both formulations became negative after two hours. The nitrogen groups on the ethylenediamine linker are responsible for the positive zeta potential of the functionalized nanoparticles at initial time points. Once this linker has eroded, the exposed free carboxylic acids on the nanoparticles contribute to the negative zeta potential. This zeta potential continued to decrease for about 12 h before it reached its saturation value of ~-18-24 mV, which is consistent with previous observations as well as with the measured zeta potential of the non-functionalized particles indicating that all the sugar from the particles has been removed (27; 30). The zeta potential of the 20:80 CPH:SA nanoparticles decreased slightly faster than that of the 20:80 CPTEG:CPH nanoparticles in accordance with the increased hydrophobicity of the 20:80 CPTEG:CPH formulation, which slowed down its erosion in aqueous solutions (37). Functionalized nanoparticles synthesized by both conventional and high throughput methods displayed nearly

213

identical zeta potentials and the kinetics of their surface potential in aqueous solution also followed similar trends, validating the high throughput method to synthesize functionalized nanoparticles.

7.4 Conclusion This work describes and validates a new automated high throughput robot as a viable tool to synthesize libraries of biodegradable polymers, of nanoparticles thereof, and of carbohydrate- functionalized nanoparticles. The automated process resulted in highly reproducible syntheses with significant time savings for the user. The high throughput method was also used to quantify the effect of adding surfactants during nanoprecipitation on nanoparticle size and protein encapsulation efficiency. With the establishment of standardized methods and protocols, future work will focus on utilizing the technique to perform high throughput screening of different nanoparticle formulations for drug delivery and cellular interactions.

7.5 Acknowledgements The authors are grateful for the assistance from Trey Hancock of OMNI International Inc.

Financial support was provided by the US Army Medical Research and Material Command

(Grant No. W81XWH-10-1-0806).

7.6 References 1. Gribbon P, Sewing A. 2005. High-throughput drug discovery: what can we expect from HTS? Drug discovery today 10:17-22

2. Abeylath SC, Ganta S, Iyer AK, Amiji M. 2011. Combinatorial-designed multifunctional polymeric nanosystems for tumor-targeted therapeutic delivery. Accounts of chemical research 44:1009-17

3. Anderson DG, Lynn DM, Langer R. 2003. Semi-automated synthesis and screening of a large library of degradable cationic polymers for gene delivery. Angewandte Chemie 42:3153-8

214

4. Stutz CC, Zhang X, Shusta EV. 2014. Combinatorial approaches for the identification of brain drug delivery targets. Current pharmaceutical design 20:1564-76

5. Lewis JS, Roche C, Zhang Y, Brusko TM, Wasserfall CH, et al. 2014. Combinatorial delivery of immunosuppressive factors to dendritic cells using dual-sized microspheres. Journal of materials chemistry. B, Materials for biology and medicine 2:2562-74

6. Petersen LK, Narasimhan B. 2008. Combinatorial design of biomaterials for drug delivery: opportunities and challenges. Expert Opin Drug Deliv 5:837-46

7. Mulet X, Kennedy DF, Conn CE, Hawley A, Drummond CJ. 2010. High throughput preparation and characterisation of amphiphilic nanostructured nanoparticulate drug delivery vehicles. International journal of pharmaceutics 395:290-7

8. Petersen LK, Sackett CK, Narasimhan B. 2010. Novel, high throughput method to study in vitro protein release from polymer nanospheres. Journal of combinatorial chemistry 12:51-6

9. Petersen LK, Sackett CK, Narasimhan B. 2010. High-throughput analysis of protein stability in polyanhydride nanoparticles. Acta biomaterialia 6:3873-81

10. Vogel BM, Cabral JT, Eidelman N, Narasimhan B, Mallapragada SK. 2005. Parallel synthesis and high throughput dissolution testing of biodegradable polyanhydride copolymers. Journal of combinatorial chemistry 7:921-8

11. Meredith JC. 2009. Advances in combinatorial and high-throughput screening of biofunctional polymers for gene delivery, tissue engineering and anti-fouling coatings. J Mater Chem 19:34-45

12. Meredith JC. 2003. A current perspective on high-throughput polymer science. J Mater Sci 38:4427-37

13. Simon CG, Jr., Lin-Gibson S. 2011. Combinatorial and high-throughput screening of biomaterials. Advanced materials 23:369-87

14. Adler AF, Petersen LK, Wilson JH, Torres MP, Thorstenson JB, et al. 2009. High throughput cell-based screening of biodegradable polyanhydride libraries. Combinatorial chemistry & high throughput screening 12:634-45

215

15. Torres MP, Determan AS, Mallapragada S, Narasimhan B. 2006. Polyanhydrides. In Encyclopedia of Chemical Processing:2247-57. Taylor and Francis. Number of 2247-57 pp.

16. Torres MP, Determan AS, Anderson GL, Mallapragada SK, Narasimhan B. 2007. Amphiphilic polyanhydrides for protein stabilization and release. Biomaterials 28:108-16

17. Mallapragada SK, Narasimhan B. 2008. Immunomodulatory biomaterials. International journal of pharmaceutics 364:265-71

18. Chavez-Santoscoy AV, Roychoudhury R, Pohl NL, Wannemuehler MJ, Narasimhan B, Ramer-Tait AE. 2012. Tailoring the immune response by targeting C-type lectin receptors on alveolar macrophages using "pathogen-like" amphiphilic polyanhydride nanoparticles. Biomaterials 33:4762-72

19. Ulery BD, Phanse Y, Sinha A, Wannemuehler MJ, Narasimhan B, Bellaire BH. 2009. Polymer chemistry influences monocytic uptake of polyanhydride nanospheres. Pharm Res 26:683-90

20. Petersen LK, Xue L, Wannemuehler MJ, Rajan K, Narasimhan B. 2009. The simultaneous effect of polymer chemistry and device geometry on the in vitro activation of murine dendritic cells. Biomaterials 30:5131-42

21. Torres MP, Vogel BM, Narasimhan B, Mallapragada SK. 2006. Synthesis and characterization of novel polyanhydrides with tailored erosion mechanisms. Journal of biomedical materials research. Part A 76:102-10

22. Conix A. 1966. Poly[1,3-bis(p-carboxyphenoxy)-propane anhydride]. Macromolecular Synthesis 2:95-8

23. Shen E, Pizsczek R, Dziadul B, Narasimhan B. 2001. Microphase separation in bioerodible copolymers for drug delivery. Biomaterials 22:201-10

24. Kipper MJ, Shen E, Determan A, Narasimhan B. 2002. Design of an injectable system based on bioerodible polyanhydride microspheres for sustained drug delivery. Biomaterials 23:4405-12

25. Song EH, Osanya AO, Petersen CA, Pohl NL. 2010. Synthesis of multivalent tuberculosis and Leishmania-associated capping carbohydrates reveals structure-dependent responses allowing immune evasion. Journal of the American Chemical Society 132:11428-30

216

26. Mereyala HB, Gurrala SR. 1998. A highly diastereoselective, practical synthesis of allyl, propargyl 2,3,4,6-tetra-O-acetyl-beta-D-gluco, beta-D-galactopyranosides and allyl, propargyl heptaacetyl-beta-D-lactosides. Carbohyd Res 307:351-4

27. Carrillo-Conde B, Song EH, Chavez-Santoscoy A, Phanse Y, Ramer-Tait AE, et al. 2011. Mannose-Functionalized "Pathogen-like" Polyanhydride Nanoparticles Target C-Type Lectin Receptors on Dendritic Cells. Molecular pharmaceutics 8:1877-86

28. Carlsen PHJ, Katsukie T, Martin VS, Sharpless KB. 1981. A Greatly Improved Procedure for Ruthenium Tetroxide Catalyzed Oxidations of Organic Compounds. Journal of Organic Chemistry 46:3936-8

29. Masuko T, Minami A, Iwasaki N, Majima T, Nishimura S, Lee YC. 2005. Carbohydrate analysis by a phenol-sulfuric acid method in microplate format. Analytical biochemistry 339:69-72

30. Goodman JT, Ramirez JEV, Boggiatto PM, Roychoudhury R, Pohl NLB, et al. 2014. Nanoparticle Chemistry and Functionalization Differentially Regulates Dendritic Cell- Nanoparticle Interactions and Triggers Dendritic Cell Maturation. Particle & Particle Systems Characterization 31:1269-80

31. Carrillo-Conde BR, Darling RJ, Seiler SJ, Ramer-Tait AE, Wannemuehler MJ, Narasimhan B. 2015. Sustained release and stabilization of therapeutic antibodies using amphiphilic polyanhydride nanoparticles. Chem Eng Sci 125:98-107

32. Petersen LK, Chavez-Santoscoy AV, Narasimhan B. 2012. Combinatorial Synthesis of and high-throughput protein release from polymer film and nanoparticle libraries. Journal of visualized experiments : JoVE

33. Haughney SL, Ross KA, Boggiatto PM, Wannemuehler MJ, Narasimhan B. 2014. Effect of nanovaccine chemistry on humoral immune response kinetics and maturation. Nanoscale 6:13770-8

34. Oyewumi MO, Kumar A, Cui ZR. 2010. Nano-microparticles as immune adjuvants: correlating particle sizes and the resultant immune responses. Expert review of vaccines 9:1095-107

35. Kanchan V, Katare YK, Panda AK. 2009. Memory antibody response from antigen loaded polymer particles and the effect of antigen release kinetics. Biomaterials 30:4763-76

217

36. Demento SL, Cui W, Criscione JM, Stern E, Tulipan J, et al. 2012. Role of sustained antigen release from nanoparticle vaccines in shaping the T cell memory phenotype. Biomaterials 33:4957-64

37. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

38. Kipper MJ, Narasimhan B. 2005. Molecular description of erosion phenomena in biodegradable polymers. Macromolecules 38:1989-99

39. Opawale FO, Burgess DJ. 1998. Influence of Interfacial Properties of Lipophilic Surfactants on Water-in-Oil Emulsion Stability. Journal of colloid and interface science 197:142-50

40. Miller R, Kragel J, Fainerman VB, Makievski AV, Grigoriev DO, et al. 2001. Encyclopedic Handbook of Emulsion Technology. New York, NY: Marcel Dekker, Inc. 731 pp.

41. Yoshioka T, Sternberg B, Florence AT. 1994. Preparation and Properties of Vesicles (Niosomes) of Sorbitan Monoesters (Span-20, Span-40, Span-60 and Span-80) and a Sorbitan Triester (Span-85). International journal of pharmaceutics 105:1-6

42. Vela-Ramirez JE, Goodman JT, Boggiatto PM, Roychoudhury R, Pohl NL, et al. 2014. Safety and Biocompatibility of Carbohydrate-Functionalized Polyanhydride Nanoparticles. The AAPS journal

218

7.7. Figures

Figure 7.1 High throughput setup. A) The robot with its three pumps (left) and its two racks for polymer synthesis (right rack) and nanoparticle fabrication (left rack). A sonicator, a stainless steel nozzle, and two fixed plastic nozzles move along a guide for filling and transferring liquid into individual tubes within a rack. B) User interface allowing the user to select a program to operate and monitor the process. C) Three syringe pumps. D) Filling the glass test tubes in rack

1. E) The rack can be easily removed and inserted into a vacuum oven enabling polymer synthesis.

219

Figure 7.2 SEM images of 1% Ova loaded nanoparticles synthesized using conventional and high throughput methods. Scale bar is 4μm.

220

Figure 7.3 Nanoparticle size is influenced by nozzle velocity during anti-solvent nanoprecipitation with the high throughput robot. Slower velocities resulted in larger nanoparticles while larger velocities decreased nanoparticle size. Scale bar is 5μm.

Figure 7.4 Cumulative mass fraction of Ova released from the nanoparticles for one month in

PBS (pH=7.4) using the microBCA assay. Error bars represent standard deviation. Experiments were performed in triplicate.

221

Figure 7.5 Zeta potential of linker-functionalized nanoparticles as a function of time in PBS

(pH=7.4). The ethylenediamine results in nanoparticles with positive zeta potential. As the particles erode, ethylenediamine is cleaved and the carboxylic acids contribute to a negative zeta potential. Error bars represent standard deviation. Experiments were performed in triplicate.

7.8 Tables Table 7.1 Characterization of polyanhydride copolymers synthesized using high throughput method. T (°C) T (°C) Copolymer Actual Molar Ratio Molecular Weight (Da) g m 100:0 CPTEG:CPH 100:0 5400 19.2 - 80:20 CPTEG:CPH 73:27 5900 20.8 - 50:50 CPTEG:CPH 45:55 8500 22.3 - 20:80 CPTEG:CPH 22:78 7600 34.7 - 0:100 CPTEG:CPH 0:100 10500 40.4 - 80:20 CPH:SA 83:17 16400 35.1 53.2 50:50 CPH:SA 48:52 12900 40 59.9 20:80 CPH:SA 16:84 12200 54.9 68 0:100 CPH:SA 0:100 9800 61.4 79.6

Table 7.2 Nanoparticle size by conventional and high throughput syntheses. 222

Conventional Synthesis High Throughput Synthesis

Span wt% 20:80 CPH:SA 20:80 CPTEG:CPH Span wt% 20:80 CPH:SA 20:80 CPTEG:CPH Diameter (nm) Diameter (nm) Diameter (nm) Diameter (nm)

0% 708 ± 322 148 ± 50 0% 359 ± 105 171 ± 53

0.2% 483 ± 102 142 ± 44 0.2% 298 ± 72 170 ± 48

1% 408 ± 119 N/A 1% 230 ± 47 N/A

Table 7.3 Protein encapsulation efficiency of nanoparticles synthesized by conventional and high throughput techniques. High Throughput Synthesis Conventional Synthesis 0 Span 20:80 CPH:SA 52.3% 0 Span 20:80 CPH:SA 71.7%

0.2% Span 20:80 CPH:SA 79.3% 0.2% Span 20:80 CPH:SA 84.3%

1% Span 20:80 CPH:SA 76.5% 1% Span 20:80 CPH:SA 99.2%

0% Span 20:80 CPTEG:CPH 98.3% 0% Span 20:80 CPTEG:CPH 74.5%

0.2% Span 20:80 CPTEG:CPH 110.8% 0.2% Span 20:80 CPTEG:CPH 108.1%

Table 7.4 Surface characterization of functionalized nanoparticles synthesized using high throughput method. Sugar Amount EDS Weight % Composition of Nanoparticle Surface Polymer composition 223 (µg/mg) Average C (%) Average N (%) Average O (%)

NF 20:80 CPH:SA - 88.5 ± 1.3 0.01 ± 0.03 11.4 ± 1.3

Glycolic Acid 20:80 CPH:SA - 83.0 ± 2.1 4.9 ± 1.4 11.8 ± 1.1 Lactose 20:80 CPH:SA 11.4 ± 2.7 81.3 ± 3.2 5.6 ± 1.4 13.2 ± 2.0 Galactose 20:80 CPH:SA 4.2 ± 3.2 80.0 ± 2.7 6.3 ± 2.1 13.7 ± 1.4

224

CHAPTER 8: COMBINATION NANOVACCINES BASED ON

HEMAGGLUTININ AND NUCLEOPROTEIN PROTECT AGAINST

INFLUENZA CHALLENGE

Jonathan T. Goodman1, Kathleen A. Ross1, Matthew Jefferson2, Metin Uz1,, Marian L. Kohut2,

Michael J. Wannemuehler3, Surya K. Mallapragada1, and Balaji Narasimhan1

To be submitted January, 2016

1 Department of Chemical and Biological Engineering, Iowa State University, Ames, IA 50011, U.S.A. 2 Department of Kinesiology, Iowa State University, Ames, IA 50011, U.S.A. 3 Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, Ames, IA 50011, U.S.A.

225

8.1 Abstract Developing more efficacious vaccines against H1N1 influenza virus would be a significant milestone that could reduce the detrimental effects of seasonal influenza epidemics and lower the risk of an influenza pandemic. In these studies, recombinant hemagglutinin (HA) and nucleoprotein (NP) antigens were encapsulated into nanovaccine formulations based on polyanhydride nanoparticles and/or pentablock copolymer hydrogels and their ability to protect animals against H1N1 infection was evaluated. Upon subcutaneous administration, all the nanovaccine formulations studied induced robust antibody titers relative to formulations containing soluble antigens (i.e., no adjuvants). The combination nanovaccine formulation, consisting of both nanoparticles and pentablock copolymer hydrogel, stimulated the strongest cell-mediated immune response six days after immunization, characterized by enhanced T cell circulation that readily produced IFN-γ and TNF-α upon stimulation. These responses correlated with desirable clinical outcomes in mice as evidenced by low viral titer in the lungs post- infection and no weight loss post-infection. In contrast, control mice that received saline lost nearly 20% of their body weight after infection. The lungs of these mice contained CD4+ and

CD8+ T cells that produced large amounts of pro-inflammatory cytokines. These results support the development of combination nanovaccine formulations against the influenza A virus.

8.2 Introduction Influenza is a potent respiratory virus that infects mammals and birds. It is responsible for an estimated 250,000-500,000 deaths each year even though there are numerous vaccines and antivirals commercially available (1). The virus is extremely successful at avoiding the immune system because of the effects of antigenic shift or antigenic drift whereby the virus undergoes point mutations or genetic recombination during its replication cycle (2). The former results in reoccurring influenza epidemics that occur every winter in the northern hemisphere (3). During

226

this time, the most susceptible individuals are the elderly, young children, and those who are immunocompromised. Most influenza infections are caused by viral strains with the H1 and H3 subtypes (4). In addition to the annual epidemics there exists the possibility of an influenza pandemic, which occurs when a new strain of the virus is transmitted to humans. There is very little immunological memory against new strain(s), resulting in high levels of mortality and a large proportion of the population being infected by the virus (5). Many of the pandemics that occurred in the 20th and 21st centuries have been attributed to new or mutated H1 influenza strains. Notable H1 influenza outbreaks include the Spanish flu pandemic of 1918, the Fort Dix outbreak in 1976, the Russian flu epidemic of 1977, and the 2009 swine flu pandemic (5; 6).

Current subunit and inactivated vaccines protect humans by inducing a large neutralizing antibody response against the hemagglutinin (HA) protein (7). HA makes up over 80% of the protein on the viral envelope and is an easy target for antibodies (8). The main function of the

HA protein is to bind to sialic acid on cells in the respiratory system and then later serve as the fusion protein to fuse the viral envelope to the cell’s endosome (8). Previous studies have shown that vaccines inducing strong cell-mediated immune responses can lessen the severity of disease and provide long-lived protection (9; 10). This is because most antibodies are directed against the head region of HA, which can vary in structure between subtype and is readily susceptible to antigenic drift (4). This greatly contrasts with epitopes that activate T cells which tend to be directed towards internal proteins that are more conserved and consistent across strains.

Specifically, dominant T cell responses tend to have epitopes originating from the matrix protein

(M1) or the nucleoprotein (NP) (11; 12). These proteins have shown little mutation over generations of viral evolution (11).

227

It is hypothesized that recombinant protein-based nanovaccines which utilize both the

HA and the NP antigens can induce protective cell-mediated and humoral immune responses.

Recombinant antigens are inherently optimal because they can be easily scaled up for mass production compared to egg-based vaccines (13-15). This reduces cost and production time making it easier for researchers to design vaccines that antigenically match circulating strains

(16). The potential drawback is that recombinant proteins are often weak immunogens and require higher doses (16). In these studies, it is demonstrated that the reduced immunogenicity of the individual proteins is overcome by formulating the HA and NP antigens with polyanhydride nanoparticles and/or pentablock copolymer hydrogels. These adjuvants/delivery systems have been shown to serve as effective vaccine adjuvants that stimulate a strong antibody-mediated immune response and a cell-mediated immune response (17-21). Both platforms have been shown to exhibit high biocompatibility, and provide a dose sparing effect where less antigen is needed to induce a potent immune response relative to solubly administered antigens (20; 22-27).

8.3. Materials and Methods

8.2.1 Monomer, polymer, hydrogel and nanoparticle synthesis 1,6-bis(p-carboxyphenoxy)hexane (CPH) diacid was synthesized using sodium hydroxide, hydroxybenzoic acid, dibromohexane, 1-methyl-2-pyrrolidinone (Sigma-Aldrich, St.

Louis, MO); sulfuric acid, and acetone (Fisher Scientific, Waltham, MA) as described previously (28). 1,8-bis(p-carboxyphenoxy)-3,6-dioxaoctane (CPTEG) diacid was synthesized using triethylene glycol, N,N-dimethylacetamide (Sigma-Aldrich); potassium carbonate, dimethyl formamide, toluene, acetonitrile, acetic acid, sulfuric acid (Fisher Scientific); and 4-p- fluorobenzonitrile (Apollo Scientific, Cheshire, UK) using previously described methods (29).

228

These diacids were copolymerized into a molar ratio of 20:80 CPTEG:CPH using a melt polycondensation reaction (30). Chemical structure, purity, and molecular weight of the copolymer were determined by dissolving the polymer into deuterated chloroform (Cambridge

Isotope Laboratories, Andover, MA) and analyzing the 1H NMR spectrum generated from a

Varian VXR 300 MHz spectrometer. Differential scanning calorimetry (DSC, TA Instruments,

New Castle, DE) was conducted to determine the copolymer glass transition temperature.

From1H NMR measurements the 20:80 CPTEG:CPH copolymer had a molecular weight of 7700 kDa, which is consistent with previous work (31; 32). Similarly NMR analysis showed that the actual molar ratio was 22:78 CPTEG:CPH, and DSC measurements indicated that the glass transition temperature was 34°C.

Monomeric hemagglutinin (HA, ~60kDA) and the nucleoprotein (NP, 57kDa) derived from Influenza A H1N1 (A/Puerto Rico/8/34) were purchased from Sino Biological Inc.

(Beijing, China). Both are recombinant proteins with the HA being produced in Human

Embryonic Kidney (HEK) 293 cells, while the NP was produced through baculovirus infection of insect cells. 1 wt.% of each lyophilized protein were encapsulated into 20:80 CPTEG:CPH nanoparticles which were synthesized using anti-solvent nanoprecipitation with a solid-oil-oil

(S/O/O) double emulsion (33). Polymer was dissolved in methylene chloride at a concentration of 20 mg/mL and both proteins were dispersed in the solution through vigorous sonication. In a

4°C cold room, the solution was poured into a -20°C pentane bath so that the ratio of solvent to anti-solvent was 1:250. Particles were collected by vacuum filtration and their morphology was analyzed with a FEI Quanta 250 scanning electron microscope (SEM, Kyoto, Japan) after a

Quorum Q150TS sputter coater (Lewes, UK) coated them with 2 nm of iridium. The diameter of

229

200 nanoparticles was measured using ImageJ image analysis software (National Institutes of

Health, Bethesda, MD).

The hydrogel was composed of 15 wt.% poly(vinyl alcohol) (PVA), 5.9 wt.% Pluronic

F127 and 4.1 wt.% PDEAEM pentablock copolymer. The pentablock copolymer was synthesized using cuprous oxide nanoparticles, a pluronic macroinitiator, N-propyl-pyridynyl methanimine (NPPM) ligand, and polyvinylpyrrolidone (DEAEM) as previously described (20).

Pluronic macroinitiator (10 g, 0.78mmol) and cuprous oxide nanoparticles (0.24 g, 1.68 mmol) were added to toluene in a 50 mL round bottom flask. The reaction flask was degassed by vacuum-argon before 3.4 mmol NPPM ligand and 2.53 mmol DEAEM monomer were added.

Oxygen was further removed with freeze-pump-thaw cycles. The reaction proceeded for 20 h at

70°C rotating at 300 rcf. Pentablock copolymer was precipitated in chilled n-heptane and collected with a Buchner funnel. To create the hydrogel, 30 wt.% PVA was heated in PBS to

80°C and was pipetted into a solution of the pentablock copolymer to form a temperature dependent hydrogel. This hydrogel was stored at 4°C until immunization. At this time, protein was dissolved in PBS and mixed into the hydrogel using vigorous sonication over ice.

8.2.2 Animals Female BALB/c mice (6-8 weeks old) were purchased from Harlan Laboratories

(Indianapolis, IN). Bedding, caging, water, and feed were all sterilized and all mice were housed in specific pathogen-free conditions. Animal procedures were approved by the Iowa State

University Institutional Animal Care and Use Committee.

8.2.3 Immunizations Mice were immunized at day 0 and then boosted with a second immunization on day 21 of the experiment. There were six treatment groups with 12 mice per group and each mouse was administered 100 μL of total volume subcutaneously at the nape of the neck. Mice from the first

230

group served as a negative control and only received sterile phosphate saline buffer (PBS). The next group of mice received a soluble dose of 10 μg each of HA and NP. The nanoparticle treatment group received 5 μg each of HA and NP encapsulated in 500 μg of 20:80 CPTEG:CPH nanoparticles with the remaining 5 μg of each protein was administered as a soluble dose. The hydrogel treatment group received 25 mg of hydrogel loaded with 10 μg each of HA and NP protein. Another group of mice were treated with a combination nanovaccine formulation containing 500 μg of nanoparticles loaded with 5 μg each of HA and NP as well as 25 mg of hydrogel containing the remaining 5 μg each of HA and NP. The last treatment group consisted of a multi-site immunization where mice were immunized with the combination adjuvant. In these animals, 50 μL of the combination adjuvant were administered at the nape of the neck and the remaining 50 μL were given at the base of the tail. Nanoparticle and hydrogel formulations were sonicated over ice for one minute to disperse any aggregates prior to immunization.

Additionally, there were six remaining mice that received 100 μL of saline. These mice were never infected and served as a naïve positive control for the other treatment groups that were challenged with the H1N1 virus.

8.2.4 T cell analysis and flow cytometry Six days following each immunization, 100 μL of blood was collected from the saphenous vein of each mouse into heparinized capillary tubes. The blood was dispensed into cell-culture media (CTCM) containing 500 units/mL heparin (Fresenius Kabi, Germany) to prevent clotting. CTCM was composed of β-mercaptoethanol (Sigma-Aldrich), heat inactivated fetal bovine serum (Atlanta Biologicals, Atlanta, GA), RPMI 1640, penicillin/streptomycin, and

L-glutamine (Mediatech, Manassas, VA). 2 mL of Ammonium-Chloride-Potassium (ACK) lysing buffer was added to the blood to lyse the red blood cells. ACK buffer consisted of ammonium chloride, potassium bicarbonate, and ethylenediaminetetraacetic acid (Fisher

231

Scientific). After 3 min of lysing, the solution was neutralized with 3 mL of CTCM media. The suspension was centrifuged for 10 min at 250 rcf at 4°C and the supernatant was poured off. The remaining cell pellet was dispersed in 250 μL of CTCM media. A small sample was used to count the cells on a Beckman Coulter Z2 counter. 5 x 105 cells were added to their own well in a round-bottom 96 well plate. The cells were stimulated with a media solution containing 50 ng/mL phorbol myristate acetate (PMA), 500 ng/mL ionomycin, and 10 μg/mL brefeldin A for 5 h in a 37°C, 5% CO2 incubator. Following incubation, the cells were fixed and lysed in their wells for 15 min at room temperature using 100 μL BD Lysing Solution (Becton-Dickinson, East

Rutherford, NJ). The cells were kept overnight in a 4°C refrigerator. In the morning, a cocktail of fluorescent antibodies was prepared in 0.5 mg/mL saponin solution and 100 μL was allowed to incubate with the cells at 4°C for 30 min. The antibody cocktail consisted of 0.1 mg/mL Rat IgG

(Sigma-Aldrich), 10 μg/mL anti-mouse CD16/32, 0.2 μg/well eFluor 450 anti-mouse CD8b,

0.125 μg/well PE anti-mouse Granzyme B (eBioscience, San Diego, CA), 0.2 μg/well PE-Cy7,

0.2 μg/well anti-mouse IFN-γ, 0.2 μg/well APC anti-mouse TNF-α, 0.2 μg/well PerCP/Cy5.5 anti-mouse CD4 (Biolegend, San Diego, CA). Following antibody staining, the cells were fixed once more with BD lysing solution, were washed in Fluorescence Activated Cell Sorter (FACS) buffer and analyzed on a BD FACSCanto flow cytometer.

At day 50, eight days after viral infection, half the mice in each treatment group were euthanized. Their lungs were extracted with half the lung being saved for histopathology and the other half processed so that homogenate could be analyzed for T cells and activation. The same antibody staining procedure used for analyzing cells in the peripheral blood was utilized for cells derived from lung homogenate. However, now 1 x 106 cells were used during incubations instead of 5 x 105 cells. In addition to stimulating the cells with a PMA/Ionomycin/Brefeldin A cocktail,

232

a separate set of cells were plated and stimulated with 64 ng/mL recombinant mouse IL-2 from

Biolegend, 4 μM nucleoprotein peptide (NP147-155), and 10 μg/mL Brefeldin A. These cells were stimulated for 6 h in a 37°C, 5% CO2 incubator. All other processing and antibody staining was identical to cells extracted from the peripheral blood.

8.2.5 Antibody titer 100 μL of blood was collected from the saphenous vein of each mouse. The serum was separated from the blood by centrifugation at 10,000 rcf for 10 min and stored at 4°C for analysis by ELISA. Each well of a flat-bottomed, high-binding 96 well plate was coated with 100 μL of

0.5 μg/mL HA and allowed to incubate overnight at 4°C. On Day 2, the wells were emptied and

2% (w/v) gelatin was added for two hours at room temperature to block the plates and reduce nonspecific binding to the adsorbed proteins. The gelatin was removed and the plates were washed with PBS-Tween (PBST). Serum was added to the wells and 1:2 dilutions were made across the plate in a solution of PBS and 1% (v/v) goat serum. The plates were incubated overnight at 4°C. On Day 3, the plates were emptied and washed with PBST. 100 μL of 1:000

Alkaline Phosphatase (AP) conjugated anti-mouse IgG H&L (Jackson ImmunoResearch, West

Grove, PA) was added to each well and incubated for 2 h. The plates were emptied and then washed with PBST before adding 100 μL of AP substrate buffer containing 1 mg/mL p- nitrophenylphosphate. Thirty minutes later the plates were read on a spectrophotometer at wavelength of 405 nm. The same procedure was repeated for specific subclasses of IgG except

AP conjugated anti-mouse IgG1 and IgG2a were used.

8.2.6 Viral challenge At 42 days following the first immunization, mice from all treatment groups were infected with a 50 μL dose of 0.5 Hemagglutinin Units (HAU) Influenza A H1N1 (A/Puerto

Rico/8/34) obtained from Dr. Yoon at Iowa State’s Veterinary Medical Research Institute. Mice

233

were anesthetized in a chamber with 3% isoflurane in 100% O2 with a flow rate of 2.5 L/min. A naïve control group of six mice were only immunized with 100 μL of saline solution and they were not challenged with the virus. All the mice were monitored twice a day for the remaining duration of the experiment.

8.2.7 Viral load quantification and clinical evaluation Three days after challenge, half of the mice were euthanized. Their lungs were extracted, put into a cryogen tube and held in a 4°C refrigerator overnight for viral load quantification the following day using quantitative reverse-transcription polymerase chain reaction (qRT-PCR) as previously described (34; 35). Viral RNA was extracted from 100μL of homogenized lung tissue using the Ambion MagMAX Viral RNA Isolation Kit (Applied Biosystems, Foster City, CA) and the KingFisher 96 magnetic particle processor (Thermo Scientific). qRT-PCR was performed on a ABI 7900HT Sequence Detection System from Applied Biosystems using the

QuantiTect Probe RT-PCR Kit from Qiagen (Hilden, Germany). Primers were at a concentration of 0.4 μM and were synthesized by Integrated DNA Technologies (Coralville, IA) using sequences from the Influenza Sequence Database. The fluorescent probe, from Applied

Biosystems, was at a concentration of 0.2 μM. Samples that were positive for the virus had a threshold count (Ct) below 38. The viral load (TCID50 equivalent/mL) was determined using a standard curve built from positive samples with known viral titer allowing Ct to be converted to

TCID50.

The remaining half of the mice was monitored for weight loss for eight days. Each mouse was weighed individually every day and the body weight was averaged across the entire treatment group. After eight days the mice were euthanized, their lungs were extracted, and half

234

the lung was fixed in formalin solution. The other half of the lung was homogenized for flow cytometry using the aforementioned methods.

8.2.8 Statistical analysis Statistical significance from the treatment groups was performed with a one-way

Analysis of Variance (ANOVA) model followed by a post-hoc Tukey’s t-test. P-values less than

0.05 were considered significant.

8.3 Results

8.3.1 Combination nanovaccine induced T cell expansion BALB/c mice were immunized subcutaneously with the treatment groups described in the Methods section at days 0 and 21. Each treatment group contained 10 μg of NP and 10 μg of

HA antigens except for the saline negative control. Six days following each immunization, 100

μL of peripheral blood was collected from each mouse for T cell proliferation.

White blood cell populations were stained for CD4, CD8, and the cytokines IFN-γ and

TNF-α. The number of cells from each sample was counted in order to assess T cell proliferation. Of the total cell count from the peripheral blood, roughly 10-15% were CD8+ T cells while approximately 30-40% were CD4+ T cells. Peripheral blood from animals immunized with nearly all of the treatment groups contained a larger number of CD4+ T cells than the blood from animals treated with saline, but the number of CD4+ T cells was significantly larger in the animals immunized with the combination nanovaccine (Figure 8.1A). In addition, peripheral blood from animals immunized with several of the treatment groups contained larger numbers of

CD8+ T cell relative to the animals that received saline (Figure 8.1B). Cells originating from the peripheral blood of mice treated with the combination nanovaccine had the largest number of

CD8+ T cells. Cells originating from mice that received the soluble antigens were also

235

statistically different compared to cells from mice treated with saline. The number of CD8+ T cells from mice treated with the nanoparticles was higher, but not statistically significant compared to CD8+ T cells from mice treated with saline.

The ability of these T cells to become activated was analyzed by performing intracellular cytokine staining on the cells upon in vitro stimulation. The CD4+ T cells from mice treated with the soluble antigens produced the largest amounts of TNF-α and IFN-γ (Figure 8.1C). The CD4+

T cells from animals that were immunized with the nanoparticles and the combination nanovaccine formulation were able to produce some TNF-α and IFN-γ, but this was not statistically significant compared to the negative control. In contrast, the number of TNF-α and

IFN-γ producing CD8+ T cells from animals immunized with the combination nanovaccine formulation was significantly greater number compared to CD8+ T cells from mice that received saline (Figure 8.1D). Finally, the number of TNF-α and IFN-γ producing CD8+ T cells from animals immunized with the combination nanovaccine delivered at multiple sites were fewer than that from animals immunized with the combination nanovaccine at a single site.

The T cell responses were reevaluated in the peripheral blood at day 27, which was six days after the second immunization. At this time point, the animals immunized with the pentablock copolymer hydrogel formulation had largest number of CD4+ T cells, while the largest number of CD8+ T cells was observed in the animals that received the soluble antigens

(Figure 8.2A & 8.2B). The peripheral blood samples from animals that received the hydrogel, nanoparticle, and soluble treatment groups showed some upregulation of TNF-α and IFN-γ in

CD4+ T cells, but these were not statistically significant from the cells in the peripheral blood of animals treated with saline (Figure 8.2C). None of the treatment groups were able to induce a

236

statistically significant proinflammatory cytokine production in the CD8+ T cell population

(Figure 8.2D).

8.3.2 Nanovaccine formulations induced robust antibody titers At days 21 and 35, peripheral blood was collected from the immunized mice and an

ELISA was performed to quantify antibody titers against the HA antigen. Total IgG as well as

IgG1 and IgG2a subclasses were analyzed. At day 21, the total IgG antibody titer was fairly low

(Figure 8.3A). Nonetheless, the animals immunized with the nanovaccine groups showed significantly greater antibody production relative to that in the animals that received the saline and soluble antigen groups. At this time point, the average IgG titer in the animals that received the soluble antigens was 350, while the animals that received the nanoparticle and hydrogel groups had titers in excess of 1100. The animals immunized with the combination nanovaccine had the highest IgG titer at nearly 3800. The animals immunized with the combination adjuvant administered at multiple sites had an average titer of 1900. Almost all of the total IgG titer was of subclass IgG1 (Figure 8.3B & 8.3C).

Two weeks after the second immunization, the antibody titers for all treatment groups increased dramatically. The animals immunized with the combination nanovaccine had the largest antibody production with an average antibody titer of nearly 1.8 x 105 (Figure 8.3A). This was a 9-fold higher than the antibody titer in the animals that received the soluble antigens.

Antibody titers from mice that were immunized with the nanoparticle and hydrogel formulations were also statistically higher compared to the titers of animals that were immunized with the soluble antigens, with the animals immunized with the hydrogel formulation inducing antibody production nearly as well as the combination nanovaccine. The titers in the mice immunized with the multi-site combination nanovaccine were statistically equivalent to that of the mice

237

immunized with the single site combination nanovaccine. At day 35, once again, most of the antibody was of IgG1 subclass but the titers in the animals immunized with the nanoparticle, hydrogel, and both combination nanovaccine formulations contained statistically significant

IgG2a components (Figure 8.3B & 8.3C).

8.3.3 Mice immunized with nanovaccine formulations had reduced viral titer upon challenge Three days post-infection, ½ of the mice in each group (6 mice/group) were euthanized and their lungs were harvested for viral quantification using real-time PCR. The lungs of all of the mice treated with saline and soluble antigens were positive for the virus. The average TCID50

8 in the lungs of mice treated with saline was 1.3 x 10 , while the average TCID50 in the lungs of mice that were immunized with the soluble antigen was nearly 1 x 107 (Figure 8.4). The lungs of mice immunized with all the other nanovaccine formulations exhibited TCID50 values that were several orders of magnitude lower. The average viral titer in the lungs of mice immunized with the nanoparticle or hydrogel formulations was of the order of 1 x 105, while the viral titer in the lungs of mice immunized with the combination nanovaccine formulation (both single and multi- site) had an average TCID50 of around 10. Significantly, the Ct value for all of the mice receiving both forms of the combination nanovaccine was above the cut-off Ct indicating that none of these mice were positive for the influenza A virus.

8.3.4 Immunized mice lost less weight following viral challenge For eight days after administration of the virus, the weight of each mouse was recorded individually and then averaged with the other mice in its treatment group. Most mice receiving antigen did not experience a statistically significant amount of weight loss (Figure 8.5). Mice treated with saline began losing weight at day 3 post-infection and continued losing weight for the duration of the experiment. On average, mice treated with saline lost about 20% of their original body weight, which was statistically significant from the weight loss in the other

238

treatment groups. Mice immunized with soluble proteins started losing weight at day 5 and reached a minimum at day 6. By day 8, these animals started to recover. At each time point, the average weight loss in the mice receiving the soluble antigens was not significantly different from the weight loss in the mice receiving the nanovaccine formulations. However, at no time point did mice receiving the nanoparticle, hydrogel, or either combination nanovaccine formulation show any weight loss. Some of these mice in fact gained weight over the eight day period and were nearly indistinguishable from naïve mice not receiving any virus.

8.3.5 T cell expansion in the lung following viral challenge Mice were euthanized on the 8th day post-infection. One million cells from lung homogenate were stimulated with an intracellular stimulation cocktail that induced nonspecific cytokine production (PMA/Ionomycin/Brefeldin A) or induced only antigen specific cytokine

+ + production (IL-2/NP147-155/Brefeldin A) from T cells within the homogenate. CD4 and CD8 T cells derived from saline- and soluble antigen-treated mice produced the largest amounts of IFN-

γ and TNF-α following nonspecific T cell stimulation (Figure 8.6A, 8.6E). There were essentially no cells positive for these cytokines in the lungs of mice that received any of the nanovaccine formulations when compared to the lungs of the uninfected naive animals.

The CD4+ T cells showed very little antigen specific cytokine production when incubated with the NP147-155 peptide as the overall number of responding cells were greatly reduced compared to those that were nonspecifically stimulated (Figure 8.6B). CD8+ cells from mice receiving saline had significant antigen specific cytokine production, which was nearly equivalent in number to the CD8+ T cells that had been nonspecifically stimulated to produce cytokines (Figure 8.6D). Finally, many of the CD8+ T cells were positive for granzyme B, particularly cells from the mice that received saline and soluble antigens (Figure 8.6E & 8.6F).

239

Relative to cells from naïve mice, there was significant granzyme B in the mice treated with saline upon both nonspecific and antigen specific T cell stimulation.

8.4 Discussion According to the Centers for Disease Control and Prevention (CDC) the best strategy to prevent influenza and control its spread is through annual vaccination (36). Two types of vaccines are currently available to consumers to meet this need. The first is the trivalent inactivated influenza vaccine (TIV) and the second is the live attenuated influenza vaccine

(LAIV) (37). In the TIV vaccine, virions from three or four different strains are inactivated for intramuscular administration (38). TIV vaccines with purified virion surface antigens are known as subunit vaccines (38). The LAIV contains the same three or four viruses used to formulate the

TIV vaccine, but it is administered intranasally as a live vaccine that can only replicate in the nasal mucosa (39). The composition of these vaccines varies across manufacturer, but most influenza vaccines are derived from viral infection of chicken eggs (40). Although these vaccines can be effective, there are significant limitations. The first is that these vaccines provide protection mostly through an antibody-mediated immune response (7; 41). A more robust cell- mediated response in conjunction with a strong antibody response could help infected individuals recover faster and provide enhanced protection (7). Additionally, there have been influenza seasons recently where vaccine shortages have existed (42; 43). Utilizing recombinant antigens that can be scaled up more easily along with incorporating adjuvants that can provide a dose- sparing effect could reduce the potential for a vaccine shortage (44). Thirdly, egg based vaccines can have residual egg protein in the vaccine, which could be hazardous to those that are hypersensitive or allergic to eggs (45). Recombinant antigens that are incorporated into biodegradable nanoparticles and/or hydrogels avoid this problem altogether.

240

The combination nanovaccines described in this work have the potential to overcome many of these limitations in the current influenza vaccines. Our data indicates that nanovaccine formulations containing recombinant HA and NP protein can effectively adjuvant the immune response and protect animals against a live influenza A virus challenge. These nanovaccine formulations effectively stimulated T cells and generated an antibody response that was several orders of magnitude greater than vaccine formulations consisting of soluble antigens. In turn, this resulted in a more favorable clinical outcome following viral challenge. The enhanced immunological response demonstrated herein reinforces the concept that nanovaccines have dose sparing capabilities. An influenza vaccine that can minimize dosage would be a significant breakthrough given that current vaccine production capacity is inadequate to combat a potential pandemic (46).

The data in Figure 8.1A and 8.1B showed that the combination nanovaccine induced significantly larger numbers of CD4+ and CD8+ T cells. The number of CD8+ T cells in the animals vaccinated with the soluble antigens was also significantly larger. The early endogenous

T cell data indicated that the soluble antigen formulation and the combination nanovaccine performed the best. This is likely because the animals immunized with the soluble antigens resulted in CD4+ T cells that produced a large amount of pro-inflammatory cytokines, which is attributed to the immediate availability of the entire amount of antigen for processing by antigen presenting cells (Figure 8.1C). Similarly, animals immunized with the combination nanovaccine resulted in CD8+ T cells that produced large amounts of pro-inflammatory cytokines (Figure

8.1D). Previous studies have shown that it can take up to a month for 20:80 CPTEG:CPH nanoparticles to release their payload and up to a week for some hydrogel formulations (15; 20).

As a result much of the antigen in the nanovaccine formulations is not available to stimulate a

241

strong T cell response immediately, which could explain why the nanoparticle and hydrogel treatment groups did not perform as well as the soluble antigen formulation. Nonetheless, it appears that combining the nanoparticles and the hydrogel together provided a strong enough signal that adjuvants the immune response to induce an increased number of T cells even though much of the HA and NP antigen is still encapsulated within the formulation.

The endogenous T cell data following the boost immunization indicated that animals immunized with the hydrogel formulation showed a strong CD4+ T cell immune response, including production of pro-inflammatory cytokines (Figure 8.2A & 8.2C). Animals immunized with the soluble antigens showed large numbers of cytokine-producing CD8+ T cells. It is interesting that the combination nanovaccine did not induce a strong T cell response in the periphery as was observed after the primary immunization. It is known that the kinetics of T cell responses peak at 5-7 days following immunization, however, T cell responses upon secondary immunizations may reach their peak at earlier times (47; 48).

The animals immunized with the nanovaccine formulations significantly outperformed their counterparts that were immunized with soluble antigens and saline in terms of their serum antibody titers. In particular, the animals immunized with the combination nanovaccine showed the highest antibody titers 35 days after the first immunization, with titers nearly reaching 1.8 x

105 using a total of only 20 μg of HA. Most of these titers were IgG1 subclass, but at day 35, animals immunized with each of the nanovaccine formulations demonstrated some class switching and also induced IgG2a specific antibodies. The nanovaccines clearly showed a dose sparing effect, which is consistent with previous studies (17; 24). Additionally, animals immunized with the hydrogel resulted in antibody titers that were nearly equal to that induced in animals immunized with the combination nanovaccine. The hydrogels are known to form an

242

antigen depot providing a sustained release of antigen, which in turn augments the immune response (20). Antigen persistence has been shown to induce better germinal center formation and long-lived plasma cells (49; 50).

Mice receiving the multi-site combination nanovaccine formulation had strong antibody titers that were statistically equivalent to titers in animals immunized with the combination nanovaccine at one site. It was hypothesized that antigen may drain to multiple lymph nodes by administering the formulation at multiple sites, thus enhancing the immune response. However, the CD4+ and CD8+ T cell responses, particularly at day six, were the lowest in the animals immunized with the multi-site combination nanovaccine in contrast with the animals immunized with the single site combination nanovaccine.

Even though the animals immunized with the multi-site combination nanovaccine formulation did not show strong T cell responses, mice receiving this formulation had an excellent clinical outcome in terms of low viral titer and no weight loss after challenge. This reinforces the concept that an efficacious antibody response is the primary means to preventing influenza infection (7). Viral titers were lowest in the lungs of mice immunized with the single site and multi-site combination nanovaccine because none of the mice in either of these groups were positive for the virus. qRT-PCR reveled that viral titers were four orders of magnitude higher in the lungs of mice receiving the nanoparticle and hydrogel formulations. This is consistent with the antibody titer data where antibody titers were slightly lower in the sera of animals immunized with the nanoparticle and hydrogel formulations. Of all the immunized animals, the cohort immunized with the soluble antigens had the highest viral titer in their lungs.

This was also the only cohort of immunized animals to exhibit any weight loss, which occurred between days 5 and 7. It has been shown that mice with viral titers of 108 or higher exhibit

243

significant weight loss (35). As anticipated, mice receiving saline had viral titers above 108 eight days after infection and showed significant weight loss relative to the other immunization groups.

These clinical symptoms and viral titers were reflected in the T cell analysis of the lungs of mice eight days after viral infection. Proinflammatory cytokine production by CD4+ and CD8+

T cells was highest in the animals that received the saline treatment. This group of animals had approximately 4-fold greater CD4+ T cells producing proinflammatory cytokines when stimulated with the PMA/Ionomycin/Brefeldin A cocktail compared to animals that received the nanovaccine formulations. Similarly, mice receiving the saline treatment had over 150-fold greater CD8+ T cells producing proinflammatory cytokines compared with mice immunized with the nanovaccines. Given that the mice immunized with the nanovaccine formulations by themselves don’t induce more cytokines than the naive animals, it is clear that these mice have effectively resolved the infection. This is true for animals that received the single site and multi- site combination nanovaccine formulations, because there was no detectable viral titer within the lungs of these mice. However, it is interesting to observe that the proinflammatory cytokine producing T cell responses in the animals immunized with the hydrogel and nanoparticle formulations was not significantly higher relative to that in naïve mice. This is because there is a detectable viral titer within the lungs of these mice. Taken together all this data suggests that the infection is waning and would likely be cleared shortly in these mice. It is evident that the saline treated mice are still dealing with the infection. It is possible that this drastic increase in cytokine production as well as granzyme B production could be induce significant histological damage within the lung tissue.

244

Cells from the lung homogenate were stimulated with the NP147-155 peptide, a known

CD8+ T cell epitope, to ascertain T cell antigen specificity. There was no statistical difference

+ between the CD4 T cell groups, which makes sense as the NP147-155 peptide should only stimulate CD8+ T cells. It is interesting to note that nearly the same number of CD8+ T cells in the lungs of mice that received the soluble antigen formulation or saline produced pro- inflammatory cytokines in the presence of the NP147-155 peptide as the PMA/Ionomycin/Brefeldin

A cocktail. This indicates that nearly all of the CD8+ T cells infiltrating the lung are antigen- specific.

8.5 Conclusions This work demonstrates that vaccine formulations containing influenza HA and NP together with a combination of polyanhydride nanoparticles and pentablock copolymer hydrogels generated a robust antibody response against HA and provided an enhanced cell-mediated immune response. Animals immunized with the combination nanovaccine generally showed the best immune responses and were protected from influenza virus challenge, but strong antibody titers and cell-mediated responses were also measured in the animals immunized with the nanoparticle and hydrogel only formulations. The animals immunized with the nanovaccine formulations (alone or in combination) were effectively protected from an H1N1 influenza virus infection compared to animals that received soluble antigens (i.e., without any nanoadjuvants).

Mice receiving the nanovaccine formulations had lower viral titers, did not experience any weight loss after infection, and had proinflammatory cytokine induction within the lung that was comparable to that in naïve unchallenged mice. Together, these results demonstrate the promise of pursuing the design and development of efficacious influenza vaccines based on these combination nanovaccines.

245

8.6 Acknowledgements The authors would like to acknowledge Dr. Shawn Rigby of the Iowa State Flow

Cytometry Facility and Dr. Kyoung-Jin Yoon for his help with the qRT-PCR and viral load quantification. Financial support was provided by The US Army Medical Research and Material

Command (Grant No. W81XWH-10-1-0806.

8.7 References 1. Fineberg HV. 2014. Pandemic preparedness and response--lessons from the H1N1 influenza of 2009. The New England journal of medicine 370:1335-42

2. Treanor J. 2004. Influenza vaccine--outmaneuvering antigenic shift and drift. The New England journal of medicine 350:218-20

3. Molinari NA, Ortega-Sanchez IR, Messonnier ML, Thompson WW, Wortley PM, et al. 2007. The annual impact of seasonal influenza in the US: measuring disease burden and costs. Vaccine 25:5086-96

4. Nunes-Alves C. 2015. Viral infection: Stemming influenza viruses. Nature reviews. Microbiology 13:599

5. Horimoto T, Kawaoka Y. 2005. Influenza: lessons from past pandemics, warnings from current incidents. Nature reviews. Microbiology 3:591-600

6. Gaydos JC, Top FH, Jr., Hodder RA, Russell PK. 2006. Swine influenza a outbreak, Fort Dix, New Jersey, 1976. Emerging infectious diseases 12:23-8

7. Hemann EA, Kang SM, Legge KL. 2013. Protective CD8 T cell-mediated immunity against influenza A virus infection following influenza virus-like particle vaccination. Journal of immunology 191:2486-94

8. Nayak DP, Balogun RA, Yamada H, Zhou ZH, Barman S. 2009. Influenza virus morphogenesis and budding. Virus research 143:147-61

9. Thomas PG, Keating R, Hulse-Post DJ, Doherty PC. 2006. Cell-mediated protection in influenza infection. Emerging infectious diseases 12:48-54

246

10. La Gruta NL, Turner SJ. 2014. T cell mediated immunity to influenza: mechanisms of viral control. Trends in immunology 35:396-402

11. Bui HH, Peters B, Assarsson E, Mbawuike I, Sette A. 2007. Ab and T cell epitopes of influenza A virus, knowledge and opportunities. Proceedings of the National Academy of Sciences of the United States of America 104:246-51

12. Chen W, Bennink JR, Morton PA, Yewdell JW. 2002. Mice deficient in perforin, CD4+ T cells, or CD28-mediated signaling maintain the typical immunodominance hierarchies of CD8+ T-cell responses to influenza virus. Journal of virology 76:10332-7

13. Sedova ES, Shcherbinin DN, Migunov AI, Smirnov Iu A, Logunov D, et al. 2012. Recombinant influenza vaccines. Acta naturae 4:17-27

14. Ross KA, Loyd H, Wu W, Huntimer L, Ahmed S, et al. 2015. Hemagglutinin-based polyanhydride nanovaccines against H5N1 influenza elicit protective virus neutralizing titers and cell-mediated immunity. International journal of nanomedicine 10:229-43

15. Ross KA, Loyd H, Wu W, Huntimer L, Wannemuehler MJ, et al. 2014. Structural and antigenic stability of H5N1 hemagglutinin trimer upon release from polyanhydride nanoparticles. Journal of biomedical materials research. Part A 102:4161-8

16. Hansson M, Nygren PA, Stahl S. 2000. Design and production of recombinant subunit vaccines. Biotechnology and applied biochemistry 32 ( Pt 2):95-107

17. Haughney SL, Ross KA, Boggiatto PM, Wannemuehler MJ, Narasimhan B. 2014. Effect of nanovaccine chemistry on humoral immune response kinetics and maturation. Nanoscale 6:13770-8

18. Huntimer L, Ross KA, Darling RJ, Winterwood N, Boggiatto PM, et al. 2014. Polyanhydride nanovaccine platform enhances antigen-specific cytotoxic T cell responses. Technology 2:171-5

19. Kipper MJ, Wilson JH, Wannemuehler MJ, Narasimhan B. 2006. Single dose vaccine based on biodegradable polyanhydride microspheres can modulate immune response mechanism. Journal of biomedical materials research. Part A 76:798-810

20. Adams JR, Haughney SL, Mallapragada SK. 2015. Effective polymer adjuvants for sustained delivery of protein subunit vaccines. Acta biomaterialia 14:104-14

247

21. Adams JR, Mallapragada SK. 2013. Novel Atom Transfer Radical Polymerization Method to Yield Copper-Free Block Copolymeric Biomaterials. Macromol Chem Phys 214:1321-5

22. Huntimer L, Ramer-Tait AE, Petersen LK, Ross KA, Walz KA, et al. 2013. Evaluation of biocompatibility and administration site reactogenicity of polyanhydride-particle-based platform for vaccine delivery. Advanced healthcare materials 2:369-78

23. Vela-Ramirez JE, Goodman JT, Boggiatto PM, Roychoudhury R, Pohl NL, et al. 2015. Safety and biocompatibility of carbohydrate-functionalized polyanhydride nanoparticles. The AAPS journal 17:256-67

24. Huntimer L, Wilson Welder JH, Ross K, Carrillo-Conde B, Pruisner L, et al. 2013. Single immunization with a suboptimal antigen dose encapsulated into polyanhydride microparticles promotes high titer and avid antibody responses. Journal of biomedical materials research. Part B, Applied biomaterials 101:91-8

25. Determan MD, Cox JP, Mallapragada SK. 2007. Drug release from pH-responsive thermogelling pentablock copolymers. Journal of biomedical materials research. Part A 81:326-33

26. Zhang B, Jia F, Fleming MQ, Mallapragada SK. 2012. Injectable self-assembled block copolymers for sustained gene and drug co-delivery: an in vitro study. International journal of pharmaceutics 427:88-96

27. Agarwal A, Unfer R, Mallapragada SK. 2007. Investigation of in vitro biocompatibility of novel pentablock copolymers for gene delivery. Journal of biomedical materials research. Part A 81:24-39

28. Conix A. 1966. Poly[1,3-bis(p-carboxyphenoxy)-propane anhydride]. Macromolecular Synthesis 2:95-8

29. Torres MP, Vogel BM, Narasimhan B, Mallapragada SK. 2006. Synthesis and characterization of novel polyanhydrides with tailored erosion mechanisms. Journal of biomedical materials research. Part A 76:102-10

30. Torres MP, Determan AS, Anderson GL, Mallapragada SK, Narasimhan B. 2007. Amphiphilic polyanhydrides for protein stabilization and release. Biomaterials 28:108-16

248

31. Torres MP, Wilson-Welder JH, Lopac SK, Phanse Y, Carrillo-Conde B, et al. 2011. Polyanhydride microparticles enhance dendritic cell antigen presentation and activation. Acta biomaterialia 7:2857-64

32. Carrillo-Conde B, Schiltz E, Yu J, Chris Minion F, Phillips GJ, et al. 2010. Encapsulation into amphiphilic polyanhydride microparticles stabilizes Yersinia pestis antigens. Acta biomaterialia 6:3110-9

33. Lopac SK, Torres MP, Wilson-Welder JH, Wannemuehler MJ, Narasimhan B. 2009. Effect of polymer chemistry and fabrication method on protein release and stability from polyanhydride microspheres. Journal of biomedical materials research. Part B, Applied biomaterials 91:938-47

34. Warren KJ, Olson MM, Thompson NJ, Cahill ML, Wyatt TA, et al. 2015. Exercise Improves Host Response to Influenza Viral Infection in Obese and Non-Obese Mice through Different Mechanisms. PloS one 10:e0129713

35. Sim YJ, Yu S, Yoon KJ, Loiacono CM, Kohut ML. 2009. Chronic exercise reduces illness severity, decreases viral load, and results in greater anti-inflammatory effects than acute exercise during influenza infection. The Journal of infectious diseases 200:1434-42

36. Fiore AE, Uyeki TM, Broder K, Finelli L, Euler GL, et al. 2010. Prevention and control of influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2010. MMWR. Recommendations and reports : Morbidity and mortality weekly report. Recommendations and reports / Centers for Disease Control 59:1-62

37. Treanor JJ, Kotloff K, Betts RF, Belshe R, Newman F, et al. 1999. Evaluation of trivalent, live, cold-adapted (CAIV-T) and inactivated (TIV) influenza vaccines in prevention of virus infection and illness following challenge of adults with wild-type influenza A (H1N1), A (H3N2), and B viruses. Vaccine 18:899-906

38. Krammer F, Palese P. 2015. Advances in the development of influenza virus vaccines. Nature reviews. Drug discovery 14:167-82

39. Carter NJ, Curran MP. 2011. Live attenuated influenza vaccine (FluMist(R); Fluenz): a review of its use in the prevention of seasonal influenza in children and adults. Drugs 71:1591-622

40. Yang LP. 2013. Recombinant trivalent influenza vaccine (flublok((R))): a review of its use in the prevention of seasonal influenza in adults. Drugs 73:1357-66

249

41. Bodewes R, Fraaij PL, Kreijtz JH, Geelhoed-Mieras MM, Fouchier RA, et al. 2012. Annual influenza vaccination affects the development of heterosubtypic immunity. Vaccine 30:7407-10

42. Desroches CM, Blendon RJ, Benson JM. 2005. Americans' responses to the 2004 influenza vaccine shortage. Health affairs 24:822-31

43. Treanor J. 2004. Weathering the influenza vaccine crisis. The New England journal of medicine 351:2037-40

44. Ulmer JB, Valley U, Rappuoli R. 2006. Vaccine manufacturing: challenges and solutions. Nature biotechnology 24:1377-83

45. Settipane RA, Siri D, Bellanti JA. 2009. Egg allergy and influenza vaccination. Allergy and asthma proceedings : the official journal of regional and state allergy societies 30:660-5

46. Auewarakul P, Kositanont U, Sornsathapornkul P, Tothong P, Kanyok R, Thongcharoen P. 2007. Antibody responses after dose-sparing intradermal influenza vaccination. Vaccine 25:659-63

47. Turnbull EL, Wong M, Wang S, Wei X, Jones NA, et al. 2009. Kinetics of expansion of epitope-specific T cell responses during primary HIV-1 infection. Journal of immunology 182:7131-45

48. Jordan KA, Wilson EH, Tait ED, Fox BA, Roos DS, et al. 2009. Kinetics and phenotype of vaccine-induced CD8+ T-cell responses to Toxoplasma gondii. Infection and immunity 77:3894-901

49. Baumjohann D, Preite S, Reboldi A, Ronchi F, Ansel KM, et al. 2013. Persistent antigen and germinal center B cells sustain T follicular helper cell responses and phenotype. Immunity 38:596-605

50. Slifka MK, Ahmed R. 1998. Long-lived plasma cells: a mechanism for maintaining persistent antibody production. Current opinion in immunology 10:252-8

250

8.8 Figures

Figure 8.1 CD4+ and CD8+ T cell expansion in the peripheral blood six days after immunization.

A) Animals immunized with the combination nanovaccine had the largest number of CD4+ T cells. B) Animals immunized with the combination nanovaccine or soluble antigens had the largest number of CD8+ T cells. C) & D) Proinflammatory cytokine production by CD4+ T cells is strongest in animals immunized with the soluble antigens and that in CD8+ T cells is greatest in the animals immunized with the combination nanovaccine. Error bars represent standard error

251

of mean. N = 12 mice per group. * represents statistical significance (p < 0.05) compared with the saline treatment group.

Figure 8.2 CD4+ and CD8+ T cell expansion in the peripheral blood on day 27, which is 6 days after the boost immunization. A) Animals immunized with the hydrogel formulation had the largest number of CD4+ T cells. B) Animals immunized with the soluble antigens had the largest number of CD8+ T cells. C) Proinflammatory cytokine production by CD4+ T cells in animals immunized with the nanoparticle and hydrogel treatment groups is high, but not statistically

252

significant compared to that in animals that received saline. D) CD8+ T cells from all the animals exhibited little cytokine production at this time point. Error bars represent standard error of mean. N = 12 mice per group. * represents statistical significance (p < 0.05) compared with the saline treatment group.

Figure 8.3 Antibody titers from immunized animals at days 21 and 35 post-immunization. A)

Nanovaccine formulations induced significant total IgG production, with the combination nanovaccine leading the way. B) Most of the antibody titers in animals immunized by the various vaccine formulations is specific to IgG1. C) Nanovaccine formulations induce statistically

253

significant class switching to IgG2a at Day 35. N = 12 mice per group. P values less than 0.05 are significant from each other. Groups that have different letters are statistically significant at that time point.

Figure 8.4 Viral titers are highest in mice treated only with saline solution and nearly undetectable in mice receiving the combination nanovaccine. N = 6 mice per group. P values less than 0.05 are significant from each other as represented by the different letters.

254

Figure 8.5 Average body weight loss in each treatment group following infection with the H1N1 virus. Mice treated with saline lost nearly 20% of their body weight eight days after infection, which mice receiving the nanovaccine formulations on average never lost weight. Saline is statistically significant from every other group starting at Day 4. N = 6 mice per group.

255

Figure 8.6 Non-specific and antigen-specific CD4+ and CD8+ cytokine production in the lungs of mice immunized with various formulations eight days after viral infection. A) CD4+ T cells from mice receiving saline exhibit the largest cytokine production upon non-specific stimulation.

256

B) None of the mice immunized with various formulations showed antigen-specific cytokine production in CD4+ T cells. C) & D) Mice receiving saline produced overwhelming amounts of pro-inflammatory cytokines in CD8+ T cells upon both non-specific and antigen specific stimulation. E) & F) The CD8+ T cells from mice receiving saline also produced granzyme B upon both non-specific and antigen specific stimulation. In contrast, there is no CD4+ or CD8+ cytokine production upon stimulation in the lungs of mice immunized with the nanovaccine formulations. N = 6 mice per group. P values less than 0.05 are significant from each other as represented by the different letters.

257

CHAPTER 9: CONCLUSIONS AND ONGOING/FUTURE WORK

9.1 Conclusions

A significant component of this thesis has been dedicated to the rational design of nanovaccines that are capable of targeting their payloads to specific receptors on antigen presenting cells (APCs). There is evidence in the literature on the use of functionalized vaccines to augment the immunogenicity of an antigen by stimulating its uptake and processing as well as triggering proinflammatory transcription factors through intracellular signaling (1-3). Chapters 4,

5, and 6 investigated a specific class of receptors, known as C-type Lectin receptors (CLRs) that are present on APCs such as dendritic cells. Nanoparticles were functionalized with di-mannose in order to target the macrophage mannose receptor. Early results conducted on primary dendritic cells in vitro suggested that these di-mannose modified nanoparticles were readily internalized by the cells, induced elevated cytokine secretion, and stimulated surface marker expression (4).

In vivo studies in Chapter 5 showed that both functionalized and non-functionalized polyanhydride nanoparticles were safe for subcutaneous and intranasal administration. There was no histological damage in kidney, liver, or lung tissue of mice upon administration of these nanoparticles and no significant upregulation of serum biomarkers that would indicate tissue damage was observed (5). The efficacy of the carbohydrate-functionalized nanoparticles was tested with additional in vivo studies in Chapter 6. In these studies, it was observed that carbohydrate-functionalized nanoparticles encapsulated with ovalbumin did not stimulate a robust CD8+ T cell response in mice and induced similar long-lived antibody titers as those induced in animals receiving non-functionalized nanoparticles.

258

Another core component of this thesis evaluated the ability of polyanhydride nanoparticles to protect mice against influenza A virus. In Chapter 8, nanoparticles were loaded with the hemagglutinin (HA) and nucleoprotein (NP) antigens. Protection against influenza is known to depend strongly on the antibody response against HA, but the NP was added in order to help stimulate a stronger cell-mediated immune response (6). Pentablock copolymer hydrogels were used as adjuvants in combination with the nanoparticles to replace the carbohydrate- functionalized nanoparticles because previous results have suggested that this combination induced synergistic in vivo immune responses against influenza (7). HA- and NP-based nanovaccine formulations consisted of nanoparticles alone, hydrogel alone, and a combination formulation in which nanoparticles were combined with hydrogels. The combination nanovaccine induced the best cell-mediated immune response as evidenced by the increased number of activated T cells. In addition, all nanovaccine formulations generated strong antibody titers. Mice treated with the nanovaccines had reduced clinical morbidity after viral infection, which was exemplified by almost no weight loss and lower viral titer in their lungs post- infection. All the mice immunized with the combination nanovaccine completely cleared the virus, while all the mice receiving saline and soluble antigens had significantly higher viral nucleic acids in their lungs.

Overall, this thesis demonstrates the enhanced properties of polyanhydride nanoparticles as nano-adjuvants that can induce both antibody- and cell-mediated immune responses against viral pathogens such as influenza A. This work also developed a new high throughput methodology for efficient synthesis and functionalization of these nanoparticles. Although the carbohydrate-functionalized nanoparticles performed similarly to the non-functionalized nanoparticles in terms of inducing in vivo adaptive immune responses, insightful information

259

was still obtained from these studies that provides some new directions for using functionalized polyanhydrides as an adjuvant/delivery platform. Targeting CLRs for drug delivery is still a nascent field. The concept of lectins recognizing specific glycan moieties was discovered in the

1980s and the structure and function of many lectins are still unknown (8). Future research will undoubtedly illuminate the targeting capabilities of this class of receptors. The in vivo influenza challenge studies in Chapter 8 showed that polyanhydride nanoparticles provide an enhanced immune response and when used in combination with the pentablock copolymer hydrogels, they provide synergistically enhanced immune responses, which significantly reduced clinical morbidity. All of these studies with the nanovaccines have motivated several ongoing research directions and future studies, which are described below.

9.2 Ongoing Research and Future Studies

There are several ongoing studies with our collaborators that are further evaluating the efficacy of polyanhydride nanoparticles as vaccine delivery vehicles, particularly against influenza A virus. These studies illustrate the potential breadth of the advantages provided by the polyanhydride nanoparticles for the development of influenza vaccines. For example, preliminary data generated together with our collaborators, Drs. Legge and Waldschmidt, at the

University of Iowa has indicated that nanovaccine formulations based on polyanhydride nanoparticles have the ability to provide heterosubtypic protection. Ongoing studies with our collaborator, Dr. Kohut in the Department of Kinesiology at Iowa State University, show that nanovaccines may induce efficacious immune responses in aged mice in addition to young adult mice. Finally, ongoing studies with our collaborator, Dr. Gourapura, at Ohio State University are evaluating whether polyanhydride nanoparticles can protect pigs from influenza viral challenge.

All of these studies are described below.

260

In our collaborative studies with researchers at the University of Iowa, 20:80

CPTEG:CPH nanoparticles were synthesized with 1 wt.% HA, 1 wt.% NP, and 1.67 wt.% CpG

ODN 1668 in the presence of 0.2 wt.% Span 80 as a surfactant. CpG is a TLR 9 agonist known for stimulating the NF-κB transcription factor triggering a release of proinflammatory cytokines and kick starting the adaptive immune response (9; 10). The NP antigen contains several conserved epitopes recognized by T cells (11). The hypothesis here is that the presence of the NP will stimulate a cell-mediated immune response and help protect the mice from heterologous challenge as these epitopes tend to be more conserved across strains (12). Collaborators at the

University of Iowa intranasally immunized BALB/c mice with 500 μg of nanoparticles containing 2.5 μg each of HA and NP along with 2.5 μg of soluble HA and NP. The mice were boosted with the same dose on day 14, and at day 46 they were challenged with either the homologous H1N1 A/PR/8/34 or the heterologous H3N2 A/HK/68 virus. All the mice receiving the nanovaccine immunization survived the homologous challenge, losing approximately 10% of their body weight before recovering (Figure 9.1A & 9.1B). In contrast, naïve control mice that were infected but not vaccinated on average lost over 25% of their body weight and after 18 days post-infection resulted in 100% death. Heterologous challenge resulted in the immunized mice losing about 20% of their body weight, and 80% of the mice survived the challenge (Figure 9.1C

& 9.1D). Again, the naïve control mice lost over 25% of their body weight with only 20% survival post-challenge. Future studies are underway to demonstrate reproducibility as well as to probe the mechanisms of the role of CD8+ T cells and the kinetics of the B cell response in inducing protection following nanovaccine immunization.

Other collaborative studies here at ISU are analyzing the ability of polyanhydride nanovaccines to induce an efficacious immune response in aged mice. It is known that the

261

strength of the immune response to a new antigen is strongly correlated with age (13). The elderly generally have a reduced adaptive immune response when immunized with a new antigen

(14). It has been shown that the elderly on average have higher basal levels of proinflammatory cytokines and need more antigen than a healthy young adult to stimulate cytokine production against that antigen (15). Our collaborative studies with Dr. Kohut analyzed the ability of polyanhydride nanoparticles to stimulate an efficacious cell-mediated immune response against ovalbumin (Ova). In these studies, 2 wt.% of Ova was encapsulated into 20:80 CPTEG:CPH nanoparticles using a solid/oil/oil double emulsion. Pentablock copolymer hydrogels with a composition of 15 wt.% poly(vinyl alcohol) (PVA), 5.9 wt.% Pluronic F127 and 4.1 wt.%

PDEAEM pentablock copolymer were prepared by the Mallapragada laboratory. 17 month-old and 2-month young mice were immunized subcutaneously at the nape of the neck with 100 μg of soluble protein or the combination nanovaccine consisting of 10 μg protein in the nanoparticles and 90 μg protein in the hydrogels. After six days, draining lymph nodes were collected and homogenized. Cells were stimulated with a PMA/Ionomycin/Brefledin A cocktail and analyzed via flow cytometry.

The combination nanovaccine induced greater numbers of CD4+ and CD8+ T cells in the draining lymph nodes of both young and aged mice, although this result was only significant in young mice (Figure 8.2A & 8.2B). In contrast, the mice receiving soluble Ova had numbers of T cells similar to saline administered mice. All formulations resulted in greater numbers of CD4+

IFN-γ+ TNF-α+ T cells in aged mice in comparison to young mice, but there were no differences between formulations of the same age group (Figure 8.2C). The enhanced immune response provided by the combination nanovaccine was statistically significant in the young mice.

Although there was not a statistically significant upregulation in aged mice, trends in the data

262

suggest that the combination vaccine could elevate the immune response in aged mice. As a result future studies will continue to optimize the combination nanovaccine formulation to produce a more efficacious immune response in aged mice. These studies will utilize the HA and

NP antigen.

Finally, ongoing studies with our collaborators at Ohio State University are analyzing the immune response of intranasally administered polyanhydride nanoparticles loaded with whole inactivated H1N2 swine antigen. The swine influenza virus is highly contagious in pigs and is of great interest to the swine industry due to the economic losses that it can potentially inflict on livestock (16). Additionally, pigs are considered “mixing vessels” where different subtypes of the influenza virus can easily recombine and be translated to humans (17; 18). As a result an efficacious influenza vaccine for swine is urgently needed. In these studies, 20:80 CPTEG:CPH nanoparticles were synthesized using a solid/oil/oil emulsion in the presence of 0.2 wt.% Span 80 surfactant encapsulating 4 wt.% of H1N2 antigen. 4-5 week old piglets were immunized with an

7 equivalent dose of 1 x 10 TCID50. Three weeks following initial immunization, the piglets were boosted and after a total of five weeks following the first immunization, the piglets were

6 intranasally and intratracheally challenged with heterologous H1N1-OH7 (6 x 10 TCID50/pig).

Compared to whole inactivated antigen administered in soluble form, the polyanhydride nanoparticle formulation stimulated similar numbers of helper T cells, cytotoxic T cells or γδ T cells within isolated peripheral blood mononuclear cells at 35 days post-vaccination (Figure

9.3A). Lung lesions from the piglets were scored at six days post-infection (Figure 9.3B). All infected pigs had significant lung lesions. The lungs of the piglets immunized with the nanoparticle formulation showed reduced lesions, but this was not statistically significant from the lesions observed in the lugs of the piglets that received the other formulations. The body

263

temperature of the piglets was monitored for six days post-infection (Figure 9.3C). With the exception of the 1st day, piglets immunized with the nanoparticle formulation maintained a body temperature that were close to the mock unchallenged piglets compared to the temperatures of the piglets that were immunized with the soluble antigen formulation. Future studies need to focus on better delivery techniques because the nanoparticles currently cannot be delivered through a misting device that would allow deep penetration of the vaccine into the nasal cavity.

Instead nanoparticles are administered using a nasal dropper. It is hypothesized that optimizing the nanoparticles for use with the misting device would induce more robust immune responses because deeper penetration of the vaccine would prevent it from being wasted.

9.3 References 1. Tacken PJ, de Vries IJ, Torensma R, Figdor CG. 2007. Dendritic-cell immunotherapy: from ex vivo loading to in vivo targeting. Nature reviews. Immunology 7:790-802

2. Pashine A, Valiante NM, Ulmer JB. 2005. Targeting the innate immune response with improved vaccine adjuvants. Nat Med 11:S63-8

3. Reddy ST, Swartz MA, Hubbell JA. 2006. Targeting dendritic cells with biomaterials: developing the next generation of vaccines. Trends in immunology 27:573-9

4. Goodman JT, Ramirez JEV, Boggiatto PM, Roychoudhury R, Pohl NLB, et al. 2014. Nanoparticle Chemistry and Functionalization Differentially Regulates Dendritic Cell- Nanoparticle Interactions and Triggers Dendritic Cell Maturation. Particle & Particle Systems Characterization 31:1269-80

5. Vela-Ramirez JE, Goodman JT, Boggiatto PM, Roychoudhury R, Pohl NL, et al. 2015. Safety and biocompatibility of carbohydrate-functionalized polyanhydride nanoparticles. The AAPS journal 17:256-67

6. Hemann EA, Kang SM, Legge KL. 2013. Protective CD8 T cell-mediated immunity against influenza A virus infection following influenza virus-like particle vaccination. Journal of immunology 191:2486-94

264

7. Adams JR, Haughney SL, Mallapragada SK. 2015. Effective polymer adjuvants for sustained delivery of protein subunit vaccines. Acta biomaterialia 14:104-14

8. Cummings R, McEver RP. Essentials of Glycobiology. New York: Cold Spring Harbor Laboratory Press

9. Latz E, Schoenemeyer A, Visintin A, Fitzgerald KA, Monks BG, et al. 2004. TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nature immunology 5:190-8

10. Bafica A, Scanga CA, Feng CG, Leifer C, Cheever A, Sher A. 2005. TLR9 regulates Th1 responses and cooperates with TLR2 in mediating optimal resistance to Mycobacterium tuberculosis. The Journal of experimental medicine 202:1715-24

12. Thomas PG, Keating R, Hulse-Post DJ, Doherty PC. 2006. Cell-mediated protection in influenza infection. Emerging infectious diseases 12:48-54

13. Weinberger B, Herndler-Brandstetter D, Schwanninger A, Weiskopf D, Grubeck- Loebenstein B. 2008. Biology of immune responses to vaccines in elderly persons. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 46:1078-84

14. Remarque EJ, van Beek WC, Ligthart GJ, Borst RJ, Nagelkerken L, et al. 1993. Improvement of the immunoglobulin subclass response to influenza vaccine in elderly nursing-home residents by the use of high-dose vaccines. Vaccine 11:649-54

15. Panda A, Qian F, Mohanty S, van Duin D, Newman FK, et al. 2010. Age-associated decrease in TLR function in primary human dendritic cells predicts influenza vaccine response. Journal of immunology 184:2518-27

16. Yassine HM, Lee CW, Gourapura R, Saif YM. 2010. Interspecies and intraspecies transmission of influenza A viruses: viral, host and environmental factors. Animal health research reviews / Conference of Research Workers in Animal Diseases 11:53-72

17. Vincent AL, Ma W, Lager KM, Janke BH, Richt JA. 2008. Swine influenza viruses a North American perspective. Advances in virus research 72:127-54

265

18. Khatri M, Dwivedi V, Krakowka S, Manickam C, Ali A, et al. 2010. Swine influenza H1N1 virus induces acute inflammatory immune responses in pig lungs: a potential animal model for human H1N1 influenza virus. Journal of virology 84:11210-8

9.4 Figures

Figure 9.1 Mice vaccinated intranasally with HA and NP loaded nanoparticles survive homologous and heterologous influenza challenge. A) Weight loss following H1N1 viral challenge of vaccinated and unvaccinated mice. B) Survival curve following H1N1 viral challenge of vaccinated and unvaccinated mice. C) Weight loss following H3N2 viral challenge

266

of vaccinated and unvaccinated mice. D) Survival curve following H3N2 viral challenge of vaccinated and unvaccinated mice. N = 5 mice per group.

Figure 9.2 Combination nanovaccine enhanced the number of activated T cells in BALB/c mice.

A) Greater numbers of CD4+ T cells are present in the peripheral blood of aged and young mice immunized with the combination nanovaccine. B) More CD8+ T cells are present in the peripheral blood of young mice immunized with the combination nanovaccine. C) CD4+ T cells in aged mice express more proinflammatory cytokines across all treatment groups in comparison to that in young mice. D) The combination nanovaccine showed higher levels of cytokine expression by CD8+ T cells even though it was not statistically significant. N = 8 mice per group.

Statistical significance determined by p values less than 0.05. * compares significance for mice

267

of equivalent age but different vaccine treatment group. # compares significance of a given treatment group across age groups.

Figure 9.3 Immune response in pigs upon immunization with the nanovaccine. A) The nanovaccine induces T cell expression similarly to that induced by soluble antigen 35 days following vaccination. B) Post-infection analysis of lung lesions indicates that the lesions in the lungs of pigs immunized with the nanovaccines performed were similar to the lesions in the

268

lungs of the mock vaccinated pigs. C) Post-infection the body temperature of pigs immunized with the nanovaccine vaccine is slightly lower than that of unvaccinated pigs and pigs receiving the soluble vaccine. N = 7-8 pigs per group.