Vaccine Adjuvants: Current Challenges and Future Approaches

Abstract

For humans, companion animals, and food producing animals, vaccination has been touted as the most successful medical intervention for the prevention of disease in the twentieth century. However, vaccination is not without problems. With the development of new and less reactogenic vaccine antigens, which take advantage of molecular recombinant technologies, also comes the need for more effective adjuvants that will facilitate the induction of adaptive immune responses. Furthermore, current vaccine adjuvants are successful at generating humoral or antibody mediated protection but many diseases currently plaguing humans and animals, such as tuberculosis and malaria, require cell mediated immunity for adequate protection. A comprehensive discussion is presented of current vaccine adjuvants, their effects on the induction of immune responses, and vaccine adjuvants that have shown promise in recent literature.

INTRODUCTION

Over the last 200 years, the use of vaccines has proven to be one of the most successful medical interventions in the reduction of disease caused by infectious agents.¹ For example, through vaccination, disease caused by the human smallpox virus was eradicated worldwide. Europe, the Western Pacific, and the United States have been declared polio-disease free and have stopped using the Sabin (oral-live) vaccine, now including the killed version (Salk vaccine) as part of the childhood vaccination schedule.² In veterinary medicine, control and eradication of diseases such as swine cholera, parvovirus-induced enteritis, distemper virus, and pseudorabies virus have all been achieved through intervention strategies employing vaccination programs.³ Indeed, vaccination has been touted as the greatest medical achievement in the 20th century.

Despite advancements and improvements in vaccine efficacy and implementation over the past several decades, infectious disease still remains the largest cause of death worldwide; unfortunately, many of these deaths occur in children and infants caused by diseases that are preventable by vaccination.^4,5 According to the World Health Organization (WHO), 14% of the global childhood mortality is caused by vaccine preventable diseases including measles, Haemophilus influenzae type b (Hib), Bordetella pertussis (whooping cough), and neonatal tetanus.⁶ Many challenges still remain with regard to fully realizing the health benefits of active immunization programs. Some of these obstacles include the development of single dose vaccines, methods to overcome the poor immunogenicity of recombinant and subunit immunogens, and the ability to rapidly and rationally develop vaccines against emerging pathogens. One promising strategy for addressing these challenges is the development of new vaccine adjuvants, or carriers that enhance the effectiveness of vaccines.

Current immunization practices often require multiple doses to achieve protective immunity. Health care workers have observed that dropout rates in vaccination programs can reach as high as 70% in some developing countries.⁷ Recent failures of the human chicken pox vaccine demonstrated that the current recommended single dose is not protective in an outbreak situation.⁸ Many of the patients recently contracting mumps in Canada could not document more than a single immunization.⁹ The WHO listed the development of single dose vaccines as number one in their “Grand Challenges” for human health in 2005.¹⁰ While not receiving the full regimen of a vaccine may significantly impact the development of protective immunity for humans, in most livestock systems, it is often impractical in terms of cost, labor and stress on the animal to immunize more than once.¹¹ Vaccination still remains a cost effective way to combat disease.¹² Prophylactic administration of an efficacious vaccine can be more cost effective than therapeutic treatment, more ecologically friendly than the use of anti-microbial agents (i.e., less chance of antibiotic resistant bacteria in the environment) and offers greater flexibility in management options. It is estimated that for each $1 spent on vaccines, $5–10 are saved in what would have been lost to disease.¹³ It is estimated that 30–50% of the antibiotics produced are used in agriculture, many at subtherapeutic levels in feeds to promote growth by suppressing bacterial growth.¹⁴ Emerging antibiotic resistance, changes in consumer acceptance of anti-microbial use in food producing animals, and high cost of treatment as compared to prevention dictates that novel biologics for preventing disease must be developed.¹⁵ Vaccination against infectious agents has greatly improved the health of humans, companion animals, and livestock species worldwide. A single dose vaccine, whether for humans or animals, would greatly increase patient compliance, thus improving the efficacy of many vaccines (i.e., a full dosing regimen received at once), and reduce the costs associated with vaccination programs.

Recent developments in both synthetic and naturally derived adjuvants suggest that single dose vaccines for a variety of pathogens may be realized in the near future. However, no single adjuvant will be effective for all vaccine applications. Developing new adjuvants for improved immunotherapy requires the development of complementary strategies that address all the complex variables involved in immune surveillance.¹⁶ Thus, before discussing recent developments in vaccine adjuvants, we briefly discuss innate and adaptive immunity and the various types of vaccines currently used to confer protective immunity.

INNATE AND ADAPTIVE IMMUNITY

Innate and adaptive immune systems work together as a complex integrated system.¹⁷ When cells from innate defenses recognize foreign structures or pathogens, a cascade of events ensues which functions to eliminate or contain the threat. The innate immune system is involved in surveillance and detection of foreign invaders and as such is a key target for activation by vaccine adjuvants. Innate immunity comprises of a variety of hematopoietic and cellular factors including the complement system, phagocytic cells, NK cells, naturally occurring antibodies, γδ T cells, and anti-microbial peptides.^18,19 The innate immune system uses relatively few molecules to recognize components of foreign invaders. These bacterial associated components were described by Janeway and Medzhitov as pathogen-associated molecular patterns (PAMPs).¹⁸ Depending on the vigor of the innate immune response, the adaptive immune response may or may not be actively engaged. In contrast to innate immunity, adaptive immunity recognizes antigen-specific epitopes via specialized cell surface receptors (antibody or T cell receptor) resulting in an antigen-specific and more directed immune response.¹⁸ It has been shown that a combination of innate immunity and prolonged presence of the pathogen-derived immunogens significantly influences the induction of a robust immune response.²⁰ To enhance immune activation, adjuvants can be tailored to specifically activate the type of immune response needed against a particular disease (antibody, cell-mediated, or mucosal immunity) without the need to suffer the consequences of an active infection.²¹

A critical innate immune cell that is involved with induction of immune responses is the dendritic cell (DC). DCs are found in all body tissues and, as such, are effectively distributed to play a central role in stimulation and regulation of adaptive immunity (cell mediated and humoral immunity).²² In the blood and tissues, DCs are in an “immature” state, capable of phagocytosis, and express low levels of costimulatory molecules as well as molecules associated with cellular migration (CCR7, DC-SIGN, and DEC-205).²³ In the basal and suprabasal epidermis, resident DCs or Langerhans’ cells are the first cells to encounter microbes or injected immunogens. These cells provide innate immune surveillance and are continually replenished form special progenitor cells that reside in the dermis.²⁴ Dendritic-like cells are also resident in the lungs where they discriminate between pathogenic and harmless inhaled particles.²⁴ In fact, pulmonary DCs are key producers of IL-10 and, as such, are suppressors of airway inflammation. Within the gut mucosa, DCs extend their pseudopodia between epithelial barriers to sample luminal contents.²⁵ Among the many different pattern recognition receptors (PRRs) on DCs, Toll-like receptors (TLRs) allow DCs to recognize specific microbial ligands (e.g., CpG DNA, lipoteichoic acid (LTA), lipopolysaccharide (LPS), flagellin).¹⁷ TLRs are type I transmembrane proteins that mediate the initial recognition of microbial components and as such are likely targets for stimulation by vaccine adjuvants.^26,27 Stimulation of TLR and other PRRs result in the activation of specific intracellular signaling pathways (e.g., MyD88-dependent and -independent) leading to activation of transcription factors (NFκB and/or AP-1) necessary for cellular migration, maturation, and antigen presentation. DCs acquire antigen by three main mechanisms: (1) phagocytosis or energy-dependent engulfment of bacteria, particulate matter or cellular debris; (2) macropinocytosis uptake of soluble antigens; (3) receptor mediated uptake triggered by mannose receptors, complement receptors, or Fc receptors. Upon activation via TLRs and/or other environmental cues, such as IL-8, DCs undergo maturation and migrate to the draining lymph node. Following maturation, DCs lose much of their phagocytic capacity while increasing surface expression of migratory and costimulatory molecules, such as MHC I/II, CD80, CD86, and CD40. This process is accompanied by migration to the draining lymph node(s). Within the lymph node, DCs continue maturation and serve as potent antigen presenting cells (APC) to naïve CD4⁺ and CD8⁺ T cells.

There are two antigenic processing pathways within DCs that lead to the major histocompatibility complex (MHC) molecules, whose function is to bind peptide fragments derived from pathogens and display them on the cell surface for T cell recognition.^28,29 Antigens taken up by DCs via phagocytosis are contained within a phagosome or early endosome. The phagosome fuses with a lysosome generating a phagolysosome. Following changes in the pH of the phagolysosome, proteolytic enzymes are activated and the antigen is degraded into small peptide fragments (9–13 amino acids in length) in order to facilitate their presentation to T cells and B cells. Antigens contained within phagolysosomes representing exogenous antigens are loaded into MHC II and then presented on the cellular surface for stimulation of CD4⁺ T cells. A diagrammatic representation of a mature DC presenting antigen via MHC II, the exogenous pathway, is shown in Figure 1a.

Figure 1.

Exogenous and endogenous antigen presentation. (a) Following engulfment, a pathogen or immunogenic protein is contained within a phagosome or endosome. Fusion of the phagosome with the lysosome creates a phagolysosome bringing together the engulfed antigens and degradative enzymes and MHC II molecules. Following proteolytic cleavage, MHC II chaperone protein (CLIP) is displaced by the peptide (9–13 amino acids), which binds within the MHC II cleft. The vesicle containing the peptide-MHC II complex (pMHC II) traffics through the cytosol, eventually fusing with the cell membrane and the pMHC II is now displayed on the cell surface. (b) For antigens gaining access to the cytosol of the cell (self-antigens, viruses, or cytosolic bacteria) proteins are degraded by cytosolic proteosomes or immune proteosomes. Degraded peptides are guided to TAP (transporter protein associated with antigen processing) and enter the endoplasmic reticulum. Subsequently, the peptides are loaded into MHC I molecules and following intracellular trafficking, are presented on the surface of the cell.

Antigens generated within the cytosol of the cell, including viral antigens, antigen from bacteria that escape into the cytosol, and many cancer antigens are presented by the endogenous pathway. Cytostolic proteins are degraded by proteosomes in the cytosol, chaperone proteins (TAP) translocate the peptide fragments into the endoplasmic reticulum where it is loaded into MHC I molecules that are subsequently transported to the cell surface for presentation to CD8⁺ T cells as shown in Figure 1b. While all nucleated cells in the body express MHC I molecules, only DCs are able to efficiently stimulate naïve CD8⁺ cells.³⁰ Antigen specific CD8⁺ T cells properly activated by DCs can directly kill infected cells, a powerful component of cell-mediated immunity. What also makes DCs excellent activators of adaptive immunity is that DCs regularly present antigen from the same source by both MHC I and MHC II pathways by phagocytosing necrotic or apoptotic cells, thus, allowing cytosolically derived antigens access to MHC II loading compartments.^30,31 Thus, DCs are not only involved in immune surveillance, but also act as a bridge between innate and adaptive immunity.

Both the effector and regulatory aspects of CMI and humoral immunity are directly affected by the induction or activation of CD4⁺ T helper cells. These CD4⁺ T cells can be further classified as Th1, Th17, Th2, or Treg.^32-34 A Th2-type immune response is characterized by the production of IL-4, IL-5, IL-10, and IL-13 and the secretion of IgG1 and IgE antibody isotypes. Th1-type responses are characterized by the production of IFN-γ and TNF-β, IgG2a antibodies and are usually associated with cell-mediated immunity including activated macrophages and delayed-type hypersensitivity.³⁵ Immune responses of the Th1-type are directed more towards intracellular pathogens and are necessary for clearance of many viruses, some bacteria (e.g., Mycobacterium tuberculosis) and anti-tumor effects, whereas a Th2-type response is generally associated with the induction of antibodies that effectively neutralize toxins, viruses, and bacterial adhesion.^36,37 Th17 responses are considered inflammatory in nature and are characterized by production of IL-17.³² These responses appear to provide protection during acute inflammatory reactions but have been associated with chronic inflammatory diseases. The role of Th17 cells in vaccinology or infectious disease has yet to be elucidated.

Induction of the appropriate immune response (humoral vs. CMI vs. regulatory) is essential for vaccine efficacy.^37,38 For example, in the BALB/c model of leishmaniasis, an immune response dominated by IL-4 and IgG1 (i.e., Th2-biased response), in comparison to a protective Th1-biased response (IFN-γ and IgG2a), does not protect nor allow these mice to clear the infection.^39-41 Furthermore, in regions of the world where tuberculosis is endemic, a large portion of the population is infected and presents with a preexisting immune response to Mycobacterium species that is usually Th2 dominant.⁴² It its hypothesized that the current tuberculosis vaccine (Bacillus Calmette-Guerin or BCG vaccine) is ineffective in preventing disease because the current BCG vaccine is unable to redirect the preexisting immune response (Treg and/or Th2) in to a protective, Th1 dominant immune response.^42,43 In veterinary medicine, the current vaccines used against feline infectious peritonitis viruses enhances humoral immunity which has been shown to exacerbate the disease, whereas a CMI response would be protective.⁴⁴

In addition to presentation of antigen to T cells, mature DCs help to shape the adaptive immune response by secretion of cytokines. Activated DCs produce the cytokines tumor necrosis factor alpha (TNF-α), which mediates acute inflammation, and a variety of interleukins, such as IL-1β, IL-6, IL-8, IL-12, and IL-10. The specific combination of cytokines released by activated DCs can influence the ensuing CD4⁺ T cell response. The bias of the immune response generated after antigen presentation can be characterized by measurement of the cytokine profiles upon induction of antigen-specific recall responses (Fig. 2). The production of cytokines by DCs is also a critical feature of efficacious immune induction. For example, DC secretion of IL-1β induces secretion of IL-2, which facilitates the maturation and proliferation of naïve antigen-specific T lymphocytes. Conversely, antigen presentation in the absence of effective costimulation (CD80/86, CD40) or cytokine secretion by DCs induces ineffective T cell activation that can result in either tolerance or anergy.³⁰

Figure 2.

Signals from DCs can influence the differentiation of naïve T cells. Stimulated, mature DCs present not only antigen in the context of MHC but also costimulatory surface molecules necessary for T cell activation. Furthermore, the type and quantity of cytokines secreted by DCs in conjunction with these costimulatory molecules can direct the naïve T cell into different effectors phenotypes. IL-12 secretion from the DC initiates a Th1 type response characterized by secretion of IFNγ. IL-4 secretion from the DC results in a Th2 type response characterized by the secretion of IL-4, IL-5, and IL-10. The cytokines secreted by DCs are induced following ligation of cellular receptors (PRRs or TLRs) and signals from the surrounding tissues (i.e., IL-8). New evidence is emerging regarding the role of DCs in activating Th17 and Treg cells.

VACCINES

The most potent (i.e., protective) and lasting immune response in a host is induced following a natural infection with the pathogenic organism. However, for many individuals, the clinical outcome of a naturally occurring infection may not be favorable because of a lack of an effective treatment for the given disease, untoward morbidity or sequelae, or high mortality. For these reasons, vaccines have been designed to mimic the immune response that would otherwise be induced by an active infection, thereby avoiding the undesirable effects of disease. To be effective, a vaccine must contain some portion of the disease-causing agent (e.g., bacteria, virus, or toxin) and may include an immune-enhancer or adjuvant. Vaccine regimens generally employ an initial dose or priming dose followed by two to three booster doses. This prime-boost strategy allows for the presentation of high quantities of immunogen in the draining lymph node at several time points. The first dose initiates immune responses that particularly involve DCs and naïve immune cells. Repeated administration of this same immunogen induces activation of not only effector cells (e.g., immunoglobulin-committed B cells and T cells) but also memory immune cells.²⁹ Upon subsequent exposure to the same immunogen, memory T and B cells provide for a secondary immune response characterized by a greater magnitude (e.g., high antibody titer) and one that occurs at a faster rate than the induction of a primary immune response.^29,45 Regardless of the type of immunogen administered in currently licensed vaccines (e.g., killed organism, subunit), the primary mechanism of protection is mediated by the generation of neutralizing antibodies as opposed to the induction of cell-mediated immunity.⁴⁶

Vaccines can be classified into three general categories: modified live, killed/inactivated, or subunit. Each has its advantages and disadvantages. A list of the current licensed vaccines for use in humans within the United States, is available on multiple websites managed by both the Department of Health and Human Services (DHHS) and the Centers for Disease Control and Prevention (CDC).^12,47-49 The information provided includes the type of immunogen used, the age at which the vaccine should be administered, and the immunization schedules as recommended/required for the United States as issued by the DHHS.

Live Vaccines

Other than a natural infection, vaccines containing modified live organisms, relative to other vaccine formulations, induce the most potent and lasting immune response in the host. Modified live vaccines generally require the fewest number of inoculations, require no adjuvants, often confer lifelong immunity, and can be delivered through the same route as the natural infection would occur.⁵⁰ The organism is able to replicate in the host, causing a mild, limited infection that stimulates the host immune response in a very similar fashion to that induced by a natural infection. Furthermore, these vaccines retain many of the natural microbial compounds that enhance immunity by activating the innate immune system.

Safe use of live vaccines requires that the organism first be attenuated, that is, the virulence capacity of the organism must be reduced. This can be achieved through repetitive passages (10–1,000 times) in a nonhuman host or in vitro. Alternatively, attenuated organisms can be developed by inducing genetic changes so that critical virulence attributes have been deleted or inactivated in the target organism. The Sabin oral polio vaccine and Flu-mist are two examples of modified-attenuated, live vaccines that are delivered along the same routes as the natural infection.⁵¹ A closely related but nonpathogenic organism can also be used if the nonpathogen and pathogen share immunoprotective epitopes. For example, Jenner observed that cowpox infection prevented smallpox, and an attenuated Ankara strain of vaccinia virus was used to vaccinate against smallpox.⁴ Likewise, attenuated Mycobacterium bovis used in the BCG vaccine is protective against disease caused by virulent M. tuberculosis.⁵⁰

The largest drawback of modified live vaccines is that they are able to replicate in the host and, thus, are capable of persistent infection, recombination and reversion to the virulent wild-type. If the host is immunocompromised, the organism may be able to persist, and an otherwise nonpathogenic strain may be able to induce disease in the absence of a competent immune system. The live organism may also be able to spread and induce disease in other nonvaccinated individuals. Through horizontal gene transfer and natural random mutation, attenuated organism may acquire or reacquire virulence genes and become capable of disease induction. The attenuated strain of poliovirus used in oral vaccines has been shown to circulate throughout a given population and occasionally revert to virulence.⁵¹ While modified live vaccines are very effective at inducing both cellular and humoral immunity, they can cause severe reactions, ranging from inflammation at the site of inoculation to systemic disease. Furthermore, many current diagnostic tests cannot distinguish between an individual who is naturally infected and an individual that received a modified live vaccine.^52-54 Effectiveness of live vaccines also requires that they be properly handled before administration. Keeping attenuated vaccines viable (i.e., proper storage) has been problematic in worldwide efforts to eradicate polio.⁵¹

Killed Vaccines

Killed or inactivated vaccines are comprised of the whole organism that has been treated with either heat or chemicals. In this way, the organism is not able to replicate in the host, yet cellular integrity of the pathogen is preserved. Dependent upon in vitro growth conditions, killed vaccines are also potent inducers of humoral immunity because most of the virulence factors and epitopes are present.⁵⁰ Killed vaccines do not carry the same risks as live vaccines; the organism cannot replicate and, therefore, cannot establish persistent infection, spread to other individuals, or revert to a virulent form.⁵⁰ These types of vaccines are generally cost effective to produce, possess a longer shelf life and are less sensitive to changes in temperature and handling when compared to modified live vaccines.⁵⁵ Some killed vaccines can be administered orally (e.g., typhoid and cholera) more closely mimicking natural infection.³⁵ Many injectable vaccines that contain killed/inactivated organisms include: polio virus (Salk injectable polio vaccine), whole-cell B. pertussis, hepatitis A virus, Yersinia pestis (causative agent of plague), and encephalitis viruses.⁵⁶

The use of killed vaccines often requires multiple doses for the induction of protective immunity. The degree of CMI induced following immunization with killed vaccines can be weak. Like modified live vaccines, killed vaccines are highly reactogenic and are associated with adverse side effects. For example, the whole cell killed pertussis vaccine can induce a high fever accompanied by severe pain, redness and swelling at the injection site due to the presence of LPS and other TLR ligands in the vaccine.^57,58

Subunit Vaccines

Subunit vaccines contain only a portion of the organism. Toxoids, inactivated bacterial toxins, were the first subunit vaccine to be employed for human use. Diphtheria (DT) and tetanus toxoids (TT) are formaldehyde-inactivated forms of the bacterial toxin that induce immune protection against the native toxin (i.e., neutralizing antibody). Other subunit vaccines currently in use include hemaglutinin-binding proteins of influenza virus and polysaccharide capsules of bacteria such as the vaccines that include conjugated forms of HiB (H. influenzae type B), pneumococcal (Streptococcus pneumoniae), and meningococcal (Neisseria meningitides) polysaccharides.^56,59 Because of the poor immunogenicity of carbohydrate immunogens, these compounds are generally conjugated to a protein in order to enhance the immunogenicity; this strategy has been specifically used when developing vaccines for infants or the elderly, HiB and pneumococcal vaccines, respectively.

Another type of subunit vaccine being developed does not include protein or other structural components of the pathogen but utilizes the DNA of the pathogen. By injecting the DNA sequence encoding a protective epitope, immunity can be induced against a specific pathogen that bears the target epitope.⁵⁶ DNA can be delivered using a viral vector with the epitope encoded on a plasmid or DNA-containing particulates to DCs.⁶⁰ Host cells then express the epitope, it is presented in the context of MHCI or II molecules, subsequently inducing strong cellular immunity.²⁹ While many DNA vaccines are still experimental, there are currently several DNA-based human vaccines in phase I, II, or III human trials, including vaccines against cytomegalovirus, Dengue virus, human immunodeficiency virus, herpes simplex virus-2, hepatitis B, and melanoma (skin cancer).⁶¹

Subunit vaccines offer several advantages including targeting the immune response to protective epitopes but retaining or deleting epitopes that can be used to differentiate “vaccinated” individuals from naturally exposed/infected individuals.⁶² Subunit vaccines may also eliminate many of the side effects and reactivity associated with modified live or killed whole organisms as they lack many of the microbial components that trigger innate immune recognition. The purified protein or other subunit components can be prepared free of LPS, CpG-DNA, or other TLR ligands that can induce an overt inflammatory response. Thus, subunit vaccines are very safe, and using new technologies, can be very cost effective to produce.

However, subunit vaccines still have many weaknesses. In general, subunit vaccines lack strong immunogenicity and require multiple doses for protection.⁶³ Poor immunogenicity also generally requires that subunit vaccines be delivered with an adjuvant or immunoenhancer (e.g., monophosphoryl lipid A—MPLA). Many of the bacterial components that trigger a more robust immune response also enhance the protective response by inducing affinity maturation of the antibody response, increasing serum antibody titers, and immunoglobulin class switching.⁶⁴ While current subunit vaccines can be formulated to induce high titer antibody responses, the induction of protective T cell responses (CD4⁺ or CD8⁺ cell-mediated immunity) are generally lacking.

Adjuvants enhance the immunogenicity of vaccine components where a live attenuated vaccine may not be desirable. Increasing numbers of immunocompromised patients, elderly populations, and infants represent a special problem to health officials as live-attenuated vaccines are not recommended in these groups. Subunit and recombinant protein vaccines are easier to produce and are generally considered safer than live vaccines, but require adjuvants to be efficacious.⁶⁵

ADJUVANTS

An adjuvant is an agent that stimulates the immune system, increasing the response to a vaccine, while not having any specific antigenic effect. Adjuvants are immunoenhancing materials that perform three major functions: (i) provide a “depot” for the antigen, creating an antigenic reservoir for slow release, (ii) facilitate targeting of the antigen to immune cells (APCs) and enhance phagocytosis, and (iii) modulate and enhance the type of immune response induced by the antigen alone (e.g., isotype switching induces Th1 vs. Th2 bias).^66-69 Adjuvants may also provide the danger signal that the immune system needs in order to respond to the antigen as it would during an active infection.²⁹

The first function, providing a depot for the immunogen, is accomplished by entrapping the antigen in a poorly metabolized, nondegrading or slowly degrading substance, or otherwise sequestering the antigen to allow for the slow clearance of the antigen from the body. Some of these types of adjuvants are discussed in more depth in other sections of the review. Table 1 shows the adjuvants currently being used in licensed vaccines. Aluminum phosphate and aluminum hydroxide, commonly referred to as alum, are the adjuvants most often used in human vaccines and the resulting gel-like matrix that alum creates a slow-release environment for the immunogen. Oil-water emulsions also work by sequestering the antigen and slowly releasing it. The classic water-in-oil emulsion, incomplete Freund’s adjuvant, is widely used in livestock vaccines, even though it has a tendency to induce granulomas at the injection sites.⁶⁵ It is not used in vaccine formulations for human use because of this tendency. Other mineral oil emulsions, such as Drakeol, Marcol, ISA-206, and ISA-25 are also used in various livestock vaccines.⁶⁵ Recently, MF59, a variation of the biodegradable oil squalene, has proven to be a potent adjuvant with a satisfactory safety record and thus, is suitable for human use.^65,70 Virosomes, virus-like particles, immunostimulatory complexes (ISCOMs), and liposomes all allow for the slow clearance of antigen by incorporating the antigen into small particles composed of stabilized lipids, phospholipids, or proteins. GlaxoSmithKline’s new class of adjuvants (AS02A, AS01B, AS04, and AS15-SB) combines stable mineral oil liposomes containing a squalene derivative, and immunostimulating MPLA.^61,71 Furthermore, antigen sequestering can be achieved by incorporating the antigen into microspheres composed of polymeric units of a biodegradable material. As the microsphere degrades, the antigen is released. Thus, many different carrier formulations provide antigen depots once injected.

Table 1.

Vaccine Adjuvants Currently Used in Licensed Vaccines^71,259-261

Humans, US	Humans, United Kingdom, and European Union	Livestock, Worldwide (General Categories)a
Aluminum hydroxide, aluminum phosphate, potassium aluminum sulfate (alum)	Aluminum hydroxide, aluminum phosphate, potassium aluminum sulfate (alum)	Aluminum hydroxide, aluminum phosphate, potassium aluminum sulfate (alum)
	Calcium phosphate	Saponin (QS-21)
	MF-59 (Squalene, in Fluad)	Oil emulsions paraffin, mineral oil, lanolin, squalene, ISA-70, Montanide (IMS)
	AS04 (liposome formulation containing MPLA and QS-21) (FENDrix, Cervaix)	Glycerin

^aMany livestock adjuvant-vaccine formulations are proprietary and their compositions have not been disclosed.

The second function of adjuvants is to enhance the immune response by targeting the antigen to immune cells, enhancing phagocytosis, and/or activating the APC. This can be accomplished by properties of the antigen, by a property of the carrier, or by inclusion of immunostimulatory molecules. Pertussis toxin binds with high affinity to epithelial cells, enhancing uptake of the vaccine.²⁹ Other toxins, cholera toxin (CT) and Escherichia coli heat-labile toxin (LT), bind selectively to M cells of the intestinal tract.³⁵ M cells efficiently translocate vaccine particles across the epithelial barrier into a region rich in APCs and lymphocytes.^29,65,72 While bacterial toxins such as CT and E. coli LT augment a strong humoral immune response, the response to the anti-toxin may overshadow the response to the conjugate antigen.²⁹ LPS, another bacterial-derived immunostimulant, is derived from the outer membrane of gram-negative bacteria such as B. pertussis. These bacterial products directly interact with the innate immune system via LPS receptors CD14 and TLR-4.⁷² Human TLRs, when triggered by LPS, stimulate the activation of NF-κB, a transcriptional activator for the production of pro-inflammatory cytokines.⁶⁵ Because humans are very sensitive to endotoxins, LPS is toxic for inclusion in many human vaccine preparations, and the majority of injectable solutions for medical use are pyrogen- or LPS-free.

These first two mechanisms of immunity are illustrated in Figure 3. Some adjuvants may interact directly with TLRs on APC (Fig. 3b), and can be derived from pathogens that display highly conserved structures (e.g., PAMPs).⁷³ As illustrated in Figure 3, an adjuvant can interact with the PAMP directly or release antigen as in the more traditional depot effect. Many adjuvants exhibit a combination of these characteristics.

Figure 3.

Recognition of antigen and PRR ligand by immature DC. An adjuvant may act as a depot, releasing both vaccine antigen and stimulatory PRR ligand over time (a) as in many a.u. or mineral oil formulations containing MDP, MPLA, or CpG. Conversely, the adjuvant may be directly recognized by the PRR (such as mannose receptor or TLRs) (b), as may be used in whole cell, killed bacterin vaccines or some polymer adjuvants.

Many biologically derived materials exhibit the third mechanism of adjuvanticity, modulation of the immune response mechanism. MPLA is a nontoxic LPS derivative obtained from Salmonella and has been shown to enhance IFN-γ production and induction of CD4⁺ T cell-mediated immunity.^35,74 MPLA has been shown to interact through TLR-4, however it is not fully dependant upon TLR-4 for its effect.^75,76 Ligation of TLR-4 and activation of TRIF transcription factors is responsible for activating both DCs and intraperitoneal macrophages resulting in T cell stimulation without induction of IL-6, IFN-γ, or other inflammatory molecules responsible for the toxic side-effects associated with LPS.⁷⁶

Cytokines, when included in a vaccine mixture, can enhance the immune response and/or induce immune deviation. In theory, the inclusion of recombinant cytokines can enhance the activation of the APC and also selectively direct the immune response. Delivery of IL-6 or IL-12 along with antigen induces elevated serum antibody titers of both IgG1 and IgG2a isotypes, including increased production of mucosally secreted IgA.^35,77 Inclusion of a plasmid encoding IL-2 in intranasal vaccines shifted the immune response to TT and CT (both dominant Th2-type antigens) to a Th1-type immune response.³⁵ The antibody response to antigen delivered by osmotic pump was greatly enhanced by the inclusion of IL-1β with the antigen.⁷⁸ The immune response to intramuscular plasmid DNA vaccination is enhanced by the inclusion of the gene sequence for GM-CSF.²⁹ Inclusion of exogenous cytokines in a vaccine mixture acts directly on the APC or T cell providing the secondary signal needed to induce immune activation. Many of these properties have led to inclusion of cytokine adjuvants in experimental vaccines that are currently in phase I and phase II clinical trials (Tab. 2).

Table 2.

Adjuvants Currently Being Tested in U.S. Human Clinical Trials^*

Adjuvant	Phase	Disease
Aluminum hydroxide	I	Influenza, SARS
	III	Leishmania
Alhydrogel	II	Anthrax, plague, Leishmania
Montanide ISA	I/II	Melanoma, solid tumors, malaria
QS-21	I	Cancer: breast, prostate, lung, HIV
	II	Melanoma (skin)
MF59	I/II	Influenza A (H9N2), avian flu (H5N1), HIV
ISCOMATRIX	II	Melanoma
MLPA, MPL and other TLR-4 ligands	I	HIV, visceral Leishmaniasis
	II	Allergy (tree pollen)
MDP and other TLR-2 ligands	I	HIV
AS15-SB, (liposomes with MPLA and QS-21)	III	Lung cancer
AS02A, AS01B	I/II	Malaria, HIV, melanoma
CpG (TLR-9 ligand)	I	Malaria, cancer: breast, melanoma
	II	Allergy: ragweed
Imiquimod (TLR-7 ligand)	I/II	Influenza, melanoma
Heat Liable Toxin (LTK63 and LT-R192G)	I	HIV, tuberculosis, E. coli (ETEC)
Diphtheria Toxin	II	Hepatitis B
IMP321 (Th1 activating peptide)	I	Hepatitis B
IL-12	I	HIV, Leishmania, melanoma
	II	Prostate cancer
IL-15	I	HIV
IL-2	I	HIV, melanoma
GM-CSF	I/II	HIV, cancer: melanoma, lung, ovarian, B cell lymphoma, Hepatitis B
Type I interferon	I	Influenza
Virus carrier: fowlpox, vaccina virus, canarypox	I/II	Cancer: solid tumors, breast, prostate
Bacterial carrier: Salmonella typhi CVD	I	HIV
PLG microparticles	I	HIV

^*Data from http:clinicaltrials.gov.

Immune modulation can be influenced by other characteristics of the adjuvant/delivery system.⁷⁹ As mentioned above, an immune response has been historically categorized as either Th1- or Th2-like. With the discovery of Th17 cells and the increasing role of antigen derived Treg cells in controlling disease, the relative simplicity of the Th1/Th2 paradigm will likely need modification. For the sake of this discussion, the Th1/Th2 paradigm provides a model and reference for understanding disease pathogenesis and host immunity. Many different factors can contribute to induction of an immunologically biased immune response including route of antigen delivery (intramuscular, subcutaneous, intranasal, oral), antigen dose, duration of antigen presentation, number or frequency of immunizations and inclusion of costimulatory molecules (e.g., LPS, exogenous cytokines) with the antigen.⁸⁰ Adjuvants can affect all of these factors in different ways, and hence the role of the vaccinologist is to use the correct adjuvant to induce a protective immune response.⁸⁰ In the mouse model of leishmaniasis, induction of Th2-biased immune responses by vaccination does not protect the mouse from infection nor does the mouse clear the parasite (i.e., cutaneous lesions develop and persist). On the other hand, induction of Th1-biased immunity was shown to prevent subsequent infection and lesion development illustrating that the Th1/Th2 bias of the immune response is important in the ability to induce protective immunity.^39-41 Furthermore, in examining the efficacy of BCG vaccination on the clinical outcome of tuberculosis, preexisting immune responses (usually Th2 dominant) need to be overcome and appropriately redirected in order for vaccines to be efficacious.^42,43 In laboratory animals, Ova-peptide (derived from hen egg ovalbumin, Ova) delivered in alum did not induce a T cell response that could be restimulated in vitro.⁸¹ Delivery of the same peptide within PLGA microspheres induced a significant in vitro proliferative response and production of IFN-γ when lymphocytes were restimulated in vitro with Ova.⁸¹ Cunningham et al.⁸² showed that they could alter the Th1/Th2 bias of the immune response to FliC flagellar antigen of Salmonella by changing the antigen delivery system. Antigens naturally delivered, on the surface of whole bacteria, induced predominantly IgG2a antibodies (Th1 response) whereas recombinant soluble or polymerized FliC induced primarily IgG1 and Th2 cytokines (IL-4).⁸²

Th1/Th2 Immune Modulation

Induction of the appropriate type of immune response is essential for development of protective immunity. Once naïve T cells have been primed and a Th1 or Th2 type of immune response has been initiated, further immunizations to that antigen using different adjuvants cannot shift the initial immune bias.^83,84 New or novel antigens are not affected by this previous vaccine induced bias.⁸³ However, it is believed that repeated immunizations that favor a Th2 immune bias create a situation of immunological memory that affects the ability of the immune system as a whole to initiate Th1 immune responses to subsequently encountered immunogens.⁸⁵

Table 3 summarizes the dominant antibody isotypes induced by some adjuvants, a reflection of Th1/Th2 biasing of an adjuvant. As illustrated by these examples, the form (e.g., particulate or soluble) of the antigen, delivery system, and route of delivery can all affect the Th1/Th2 bias of a subsequent immune response to a vaccine, and the type of immune response (cell-mediated or humoral) that will be protective varies with the disease in question. Antigen, adjuvants, and delivery systems need to be chosen with care to obtain the most protective response. Current licensed vaccines for the most part are lacking in their ability to induce Th1 type immune responses without also generating undesirable toxic side-effects such as the severe inflammation associated with whole-cell pertussis vaccines.³⁶ While traditional alum-based vaccines initiate the Th2 response,^65,86 a Th1 response may be more effective for preventing some diseases.⁸⁷ Alum is still widely used in veterinary vaccines, but is frequently associated with the induction of tissue granulomas and subsequent carcass losses.⁸⁸ Oil-based liposomes are capable of inducing a strong Th1 response, but are also associated with adverse tissue reactivity, granuloma formation, and subsequent carcass loss.^89,90

Table 3.

Antibody Isotype Bias Induced in Laboratory Animals by Administration of Immunogens in Various Adjuvants

Adjuvant	Examples	Basic Characteristics	Dominant Antibody Isotype	Refs.
Inorganic	Aluminum hydroxide, aluminum phosphate, and calcium phosphate	Hydrogel emulsion—creates depot effect, enhance macrophages maturation	IgG1, IgE	36,96,97,99
Oil emulsions	Mineral oil (i.e., Freunds incomplete)	Basic water-in-oil emulsion with long standing record as research gold standard	IgG1	36,63,108
	MF59	Blend of muramyl tripeptide, squalene, polyoxyethylene, sorbitan monooelate, and sorbitan trioleate	IgG2a	36,74
	QS-21	Purified saponin from Quillaja saponica, used to stabilized lipid emulsions	IgG2a	35,110
	Montamide ISA-51 and ISA-720	‘Ready to use’ oil for water-in-oil emulsion	IgG1	62,66,202
Isocoms	ISCOMATRIX	Complex of saponins and lipids	IgG1	66,110
Microbial derived	Monophosphoryl lipid A (MPLA)	Detoxified TLR-4 ligand	IgG1 and IgG2a/c	75,76,107
	Macrophage activating protein-2	TLR-2 ligand from Mycoplasma spp, and purified derivatives	IgG2a	123,124
	Virosomes	Stabilized lipid complexes containing viral proteins such as influenza hemaglutinin	IgG2a, IgA	36,103,116
	LT/CT	Modified bacterial toxins for mucosal adherence heat-liable enterotoxin and cholera toxin	LT: IgG1, IgG2a, and IgA	36,141
			CT: IgG1
	CpG	Nonmethylated bacterial DNA, a TLR-9 ligand	IgG2a	36,64,129
Cytokines as adjuvants	IL-1	Pro-inflammatory cytokine	IgG2a, IgA	35,66
	IL-2	Lymphoproliferative cytokine	IgG2a	35
	IL-12	Pro-inflammatory cytokine	IgG2a, IgA	144
	IL-6	Anti-inflammatory cytokine	IgG1, IgA	145
Natural polymers	Polysaccharides	Coating or emulsified with solid antigen	IgG1 or IgG2a depending on route	36
Synthetic polymers	Polyanhydrides	Antigen and immunostimulators emulsified into biodegradable particles ranging from 50 μm to 20 nm	Variable depending on inclusion of immunostimulants and polymer chemistry	36,148,153
	Polyesters	Antigen and immunostimulators emulsified into biodegradable particles ranging from 50 μm to 20 nm	Variable depending on immunostimulants and antigen incorporated	36,68,189,197

In the United States, the only adjuvant currently approved for use in humans is alum. However, in England and other European Union countries, MF59 is also used. MF59 is based on a biodegradable plant oil emulsion containing muramyl tripeptide (MTP).⁹¹ Highly purified MTP is a synthetic component similar to that found in mycobacterial cell walls and MTP retains immunostimulatory properties while eliminating much of the toxic effects associated with the whole bacterium.⁷⁴ MF59 is used in the H5N1 bird flu vaccine developed by Novartis. MF59 was chosen for dose-sparing benefits (i.e., less immunogen needed) and is recommended in elderly (65 and older) including those with underlying chronic conditions such as diabetes.^92,93

Vaccine adjuvants straddle a fine line between tissue toxicity and efficacy. Multiple studies in livestock species have shown that greater immunogenicity is achieved when adjuvants causing severe tissue reactivity were used. Greater antibody titers were observed in swine vaccinated with bacterins prepared with a paraffin oil or lecithin (>20%) adjuvant; however, these adjuvants are highly irritating leading to severe diffuse granulomatous tissue at the injection site with multiple foci of necrosis.⁹⁴ While adjuvants containing lower amounts (5–10%) of lecithin-based oil or alum induce less tissue irritation, the corresponding antibody titers were also much lower.⁹⁴ Vaccine adjuvants for veterinary medicine have many of the same concerns as adjuvants used for human medicine. Tissue irritation, granuloma formation, and abscess formation at the injection site are undesirable from an animal welfare viewpoint, but also can be costly to the producer due to carcass losses at time of slaughter.⁹⁴

Another consideration for the development of new adjuvants is for the induction of mucosal immunity. With few exceptions (C. tetani, rabies virus, and other insect vector borne pathogens), most pathogens enter the host via mucosal surfaces (e.g., upper respiratory, gastrointestinal, vaginal, or urinary tracts). Induction of mucosal antibody (i.e., secretory IgA) by appropriate delivery of the antigen to the mucosal associated lymphoid tissue (MALT) is the most effective way to neutralize these pathogens or their secreted toxins.⁷²

ALUM ADJUVANTS

Salts of aluminum hydroxide or aluminum phosphate, commonly referred to as alum, have long been used in vaccines and have an extensive safety record. Alum was first used as an adjuvant in 1926.⁹⁵ Until recently, it was the only adjuvant approved for use in humans.^95,96 Gels of aluminum phosphate are commercially available for clinical use and generate consistent, predictable results.⁹⁷ Alum-based vaccines are prepared by suspending the antigen in a phosphate buffered solution and allowing the antigen to adsorb to the aluminum hydrogel.⁹⁷ The amount of antigen that adsorbs onto alum depends upon the forces within the antigen, and between the antigen and the alum, including hydrophobic interactions, van der Waals forces, ionic charges, and hydrogen bonding. The typical quantity of alum in a human vaccine dose is 0.5 mg, while the upper allowable limit established by the U.S. Food and Drug Administration (FDA) and WHO is 1.25 mg per injection.⁹⁷ Alum has proven safe for routine use in children, and enhances the production of antibody to protein toxoids and polysaccharide vaccines.⁹⁷ Alum has a synergistic effect when combined with other adjuvants and can enhance the adjuvant properties of liposomes, QS-21, MPLA, and CpG.⁹⁷ However, alum is not ideal for small peptide vaccines or for use with recombinant proteins due to their inherent low immunogenicity.^35,80,98

Recently, the use of alum in vaccines has come under scrutiny. Alum has been occasionally associated with severe tissue reactions such as erythema, subcutaneous nodules, granulomas, and has been thought to induce hypersensitivity and macrophagic myofasciitis.^96,99,100 It is well established that alum-based vaccines induce IgE and IL-4, which are associated with allergy and type IV immediate hypersensitivity.⁹⁶ While alum is effective at inducing strong humoral immunity, alum-based vaccines generally fail to induce cell-mediated immune responses, such as cytotoxic T cells or delayed type hypersensitivity.³⁶ Alum enhances a strongly biased Th2 immune response in animal models.⁹⁹ Alum-based vaccines have other drawbacks besides the immune bias. Alum, because it is a semi-particulate hydrogel, cannot be lyophilized or frozen,¹⁰¹ thus limiting shelf life and storage conditions. Because the mode of action of alum includes the formation of antigenic deposits at the site of injection, alum is not suitable for oral or intranasal immunization.^35,80,102 Finally, alum proved to be ineffective when used in conjunction with DNA-based vaccines.¹⁰³

The mechanism of adjuvanticity for alum has been traditionally thought of as providing an antigenic depot in the tissue. The evidence of the depot effect, or delayed antigen release, of alum adjuvants was established by White in 1967 and Harris in 1935, by inducing immunity in a second animal by implanting granulomatous tissue that had developed as a result of immunizing the donor animal with an alum-based vaccine.⁹⁷ Alum particles have been observed at the site of injection up to a year after immunization.⁹⁷ Alum-precipitated antigens are somewhat particulate, and therefore, more readily ingested by phagocytes.¹⁰⁴ Macrophages recovered from muscle tissue following injection of an alum-based vaccine and macrophages cultured in vitro in the presence of alum show persistence of crystalline inclusions.⁹⁹ Alhydrogel and Adju-phos, commercially available prepared alum gels, produce particles roughly 3–4.5 μm in size.⁹⁷

Excess alum in a vaccine mixture enhances the adjuvant effect, however alum is slightly cytotoxic to macrophages.⁹⁷ Recent studies with cultured macrophages showed that aluminum hydroxide induces a distinct maturation pattern characterized by the expression of surface markers that resemble those found on mature myeloid DCs (HLA-DR^hish/CD86^high/CD83⁺/CD1a⁻/CD14⁻) endowing them with the ability to enhance activation of CD4⁺ T cells.⁹⁹ Other recent studies have shown that alum may facilitate this DC maturation by inducing the release of uric acid crystals.¹⁰⁵ Uric acid crystals are an endogenous ligand for TLR-2.^43,106 Further evidence for TLR activation was shown by a diminished response to antigens in alum injected into MyD88-deficient mice.¹⁰⁵

ADJUVANT ACTIVITY OF CALCIUM PHOSPHATE

Calcium phosphate has been used for many years as the adjuvant in childhood diphtheria–tetanus–pertussis (DTP) vaccine formulations in France.^96,97 Furthermore, calcium phosphate is a normal body constituent and is readily absorbed.^96,97 In contrast to aluminum phosphate, calcium phosphate does not induce IgE production in animals or humans.^96,97 Because of this property, the most common use of calcium phosphate is the delivery of allergens in desensitization therapy for allergic patients.^96,97 In laboratory animals (e.g., mice and guinea pigs), calcium phosphate elicits a lower antibody response than alum-based preparations, however, the opposite is true in humans.⁹⁶ Using calcium phosphate-based vaccines, children and pregnant women developed higher neutralizing antibodies than those receiving an aluminum phosphate-based vaccine.⁹⁷ The mode of action is thought to be the same as for alum compounds, functioning to create a depot for the immunogen and facilitating the uptake of the particulate antigen by APCs.⁹⁷

FREUND′S COMPLETE ADJUVANT AND FREUND′S INCOMPLETE ADJUVANT

Freund’s complete and incomplete adjuvants (CFA and IFA, respectively) are the standard classical adjuvants to which all other adjuvants are compared.⁸⁰ This very potent adjuvant system is comprised of a water-in-mineral oil emulsion with the emulsifier mannide monooleate.¹⁰⁷ Freund’s complete adjuvant also contains heat-killed M. tuberculosis whereas IFA contains only the mineral oil emulsion and emulsifier.^29,63,107 Classically, proteinaceous antigens administered in CFA induce a very strong immune response, including cell-mediated responses, whereas immunogenic proteins administered intraperitoneally in IFA were thought to induce tolerance.^83,108 Advances in both knowledge of the immune system (induction of tolerance and Th2 responses) and methodology in measuring immune responses have shown that administration of antigens in IFA actually induces a Th2 response. This response is characterized by the induction of memory T cells that traffic to the spleen, rather than the draining lymph nodes.¹⁰⁸ In addition, the cytokine response produced by these cells is small in quantity (as compared to Th1 cytokines in a lymph node) and may be below the limits of detection.^83,109 The presence of the mycobacterial products in CFA provide a potent danger signal and induces costimulatory signals necessary for induction of Th1-type cytokines. Thus, the resultant immune responses induced by CFA and IFA provide the basis for the differential Th1/Th2 skewing of the immune response (i.e., immune deviation) observed when these two similar adjuvants are employed in a vaccine.¹⁰⁸ Complete Freund’s adjuvant is capable of inducing high antibody titers and long lasting T cell responses, but is so reactogenic that its use even in laboratory animals is discouraged.¹⁰⁷ The immune enhancing mechanisms of these adjuvants, the delayed release of antigen, slower antigen clearance, and targeting of the antigen to APCs is due to the mineral oil emulsion.²⁹ Variations on mineral oil emulsion vaccine adjuvants are marketed by Chiron and Norvarits as Montamide ISA-51 and ISA-720.⁹²

MF59 OIL-EMULSION ADJUVANTS

Introduced in Europe in 1997, MF59 is an oil-in-water microemulsion that includes squalene (derived from biodegradable plant oil), Polysor-bate 80, and Span 85 (stabilizers) and small amount of MTP, a novel synthetic component derived from mycobacterial cell walls.^63,74 In clinical trials, the MTPs proved to be too toxic and have been excluded from current formulations.^63,70,74 MF59 has been shown to stimulate a strong Th2 biased immune response to a large number of antigens and may be more suitable for subunit vaccines than alum.³⁶ MF59-based vaccines that have incorporated recombinant antigens induce high titer antibody responses and T cell proliferative responses.⁷⁴ Combination of MF59 with influenza subunits enhanced the immune response of elderly patients over that obtained using other adjuvants and is being evaluated for use in children.³⁶ MF59 does not induce Th1-type immunity (e.g., IFN-γ) and, therefore, may not be suitable for vaccines where cell-mediated immunity is needed for protection.⁷⁴ The mechanism of adjuvanticity for MF59 appears to be in directing delivery of the immunogen to APCs.⁷⁴ Studies with MF59 have shown that macrophages, but not DCs, are the main cell type involved in clearing the oil depot from tissue, and DCs are the key APCs within the T cell zones of the lymph node.³⁶ It was proposed that following uptake, adjuvant-induced cell death allowed for the transfer of the antigen from the macrophage to the DC for T cell induction (i.e., cross-presentation).³⁶ Another observation that arose during the development of MF59 is that there is a difference in emulsion particle size and the resulting immune response in different animal species. Small laboratory animals (mice, guinea pigs, and rabbits) develop high antibody titers following immunization with oil emulsion formulas regardless of particle size. However, nonhuman primates (baboons, chimpanzees) and goats require stable, small droplet emulsions for optimal antibody induction.⁷⁰ The key lesson here is that not all animal species respond equally to an adjuvant and testing in both large and small animals may be necessary to ensure applicability of a novel adjuvant. Mineral oil emulsions of various compositions are widely used in veterinary adjuvants, and as their safety record is improved, they are also being developed for human use.^61,62,65

IMMUNOSTIMULATING COMPLEXES (ISCOMs)

ISCOMs were first described in 1984 by Morein et al.¹¹⁰ Cholesterol mixed with plant-derived saponins under controlled conditions creates 40 nm cage-like particles referred to as immunostimulating complexes. These synthetic adjuvants are based on the concept of packaging the antigen into micro/nanoparticles or micelles, where the particle size is a crucial determinant of efficient uptake. Many different plant-derived saponins have been investigated for adjuvant activity including saponins derived from Bupleurum chinense, Glycyrrhiza uralensis, Quillaja brasiliensis, and Quillaja saponaria.^110-114 These heterogeneous compounds stabilize the lipid–cholesterol structure while adding immunostimulatory properties. However, these compounds are also generally hemolytic and their tissue-reactive toxic nature has plagued development. While saponins have been used in veterinary vaccines for many years, a balance between potency and adverse reactions will need to be achieved for widespread acceptance in human vaccines.^64,107 A detoxified saponin derivative, QS-21, has exhibited marked decrease in toxicity while maintaining the strong immunoenhancing properties.^35,110 This adjuvant has been shown to induce a strong Th1 immune response (CTL, IL-2, IFN-γ, and IgG2a) because of the lipid–cholesterol makeup. Like virosomes, ISCOMs have the ability to fuse with cellular membranes and to deliver the immunogen into the cytosol of the target cell. This results in the endogenous processing and presentation of the immunogenic peptide via MHC I.^35,64,110 This property also makes ISCOMs good vehicles for intracellular delivery of DNA-based vaccines.¹¹⁰ To increase antigenic loading of ISCOMs, affinity tags or aliphatic regions can be incorporated into recombinant proteins for higher efficiencies of incorporation into ISCOM membranes; alternatively, chelating agents (e.g., Cu⁺⁺) can be used to increase antigen binding.¹⁰⁷

VIROSOMES AND VIRUS-LIKE PARTICLES

Virosomes are particles of stabilized membrane lipids and functional viral fusion proteins that can be used to deliver vaccine antigens.^103,115 While theoretically a wide number of virus fusion proteins could be used, the majority of virosomes utilize the hemaglutinin (HA) and neuraminidase (NA) from influenza virus.¹⁰³ Virus-like particles are the spontaneous assembly of viral coat proteins lacking in viral genetic material.¹¹⁶ Virosomes and virus-like particles can be generated by either inserting the viral fusion proteins and antigen into preformed small phospholipid vesicles (liposomes) or by separation and reconstitution of viral envelopes with the vaccine antigen.¹⁰³ These particles retain the receptor binding capacity and mimic infectivity of native viruses without the risks associated with attenuated viruses and are capable of delivering vaccine antigens directly into the cytosol of the target cell.¹⁰³ This allows for induction of both humoral and cell mediated immunity because some of the virosome-delivered antigens have the potential to be presented via MHC II following endosomal processing, and virosomes that escape into the cytosol will allow for antigenic presentation via the MHC I pathway.¹⁰³ This type of delivery system has been shown to greatly enhance production of serum IgG and IgA at mucosal surfaces.³⁵ A synergistic effect is observed when other adjuvants or immunomodulators are included, such as heat-labile toxin of E. coli.³⁵ Virosomes and other virus-like particles are proving efficient for intranasal or mucosal delivery of many types of proteinaceous antigens (i.e., viral coat proteins) or DNA-based vaccines.³⁵

LIPOPOLYSACCHARIDE (LPS)

Many antigenic preparations, particularly recombinantly derived antigens, contain residual amounts of bacterial LPS and other TLR ligands that may provide adjuvant activity.¹¹⁷ LPS is known to stimulate a variety of cells to produce cytokines and chemokines that control DC trafficking and maturation.¹¹⁸ An unusual feature of its adjuvanticity is that LPS can be delivered at a different site and a different time than the antigen and still enhance the immune response to the given antigen. But despite its potency, LPS has been used only as an experimental adjuvant in animal studies due to its toxicity and pyrogenicity in humans. Chemically modified forms of its active component such as MPLA, have been shown to possess many of the adjuvant properties of LPS but without the associated toxicity.

MONOPHOSPHORYL LIPID A (MPLA)

Gram-negative bacterial extracts have strong immunopotentiating effects, however they are too toxic for routine use in human vaccines. Most of the immunostimulatory or toxic effects are derived from the lipid A portion of LPS, which is located in the outer-membrane of gram-negative bacteria.¹⁰⁷ Further analysis showed that by removing a phosphate group, sugar moiety, and an ester-linked fatty acid group the toxicity could be reduced 100- to 1000-fold, while still retaining the immunostimulatory function.¹⁰⁷ MPLA, the resulting molecule, was derived from Salmonella minnesota.³⁵ Similar to LPS, MPLA interacts with TLR-4 on APCs, although immune enhancement is observed in the absence of TLR-4.^75,76 MPLA initiates signaling through TRIF transcriptional activation rather than NF-κB, which induces many pro-inflammatory cytokines associated with the toxic effects of LPS.⁷⁶ Equivalent T-cell mediated responses were observed in mice immunized with Ova adjuvanted with LPS or MPLA indicating that the mechanism of TLR-4 signaling (TRIF vs. NF-κB), and not the magnitude of the response, was responsible for the reduction in toxicity.⁷⁶ Binding of MPLA to TLR-4 initiates the synthesis of IL-1β, IL-12, and IFN-γ, all of which are necessary for DC maturation, migration, and initiation of the T cell response.^35,119 In animal studies, MPLA induced a strong systemic Th1 type immune response, including cytotoxic T lymphocytes (CTLs).³⁵ Furthermore, MPLA was shown to enhance the production of complement fixing antibodies and increased production of secretory IgA.³⁵ While MPLA enhanced the resulting immune response to a given antigen in comparison to the immune response to the soluble antigen alone, MPLA is more effective when combined with other adjuvants or delivery systems such as alum, QS-21 (Quil A) and polymeric microspheres, or other adjuvants that provide a depot effect.³⁵ Several vaccine formulations using MPLA as an adjuvant are in clinical trials for humans and livestock species.^61,65

TLR-2 LIGANDS

Since the discovery of TLRs as a key sensing and signaling mechanism for APCs, efforts have been made to exploit TLRs as receptors for vaccine adjuvants.^120,121 Many different derivatives of gram positive cell wall components have all been found to trigger immune activation through TLR-2. OspA of Borrelia burgdorferi was used in the vaccine against Lyme disease.¹²² Muramyl dipeptide (MDP) has been synthesized from several gram positive bacteria including several Mycobacterium species, Corynebacterium granulosum, B. pertussis, and N. meningitides. MDP derivatives have been shown to induce dichotomous effects on the immune system. When delivered in soluble delivery systems, MDP enhances humoral immunity; when delivered in liposomes, MDP enhances CMI.¹¹⁶ Addition of MDP to a vaccine formulation acts synergistically with mineral oil and alum carriers, enhancing the CMI response.¹⁰⁷ Macrophage activating lipopeptide-2 (MALP-2) is another TLR-2 targeted ligand showing promise as a vaccine adjuvant. MALP-2 is an agonist of the TLR-2–TLR-6 heterodimer from Mycoplasma fermentans and has been shown to activate APCs via MyD88 signaling and activation of NFκB transcription factor.¹²³ TLR-2 and TLR-6 are also present on B cells.¹²³ Studies in mice lacking either B or T cells showed that MALP-2 activated B cells in a T cell-independent manner but enhanced T cell function via a B-cell dependant mechanism.¹²³ Pam2Cys is a synthetic compound with structural similarity to MALP-2 and has been shown to enhance the CMI and humoral response in an experimental vaccine for Listeria monocytogenes and an intranasal administration of an influenza vaccine in mice.¹²⁴ ESAT-6, a protein derived from the cell wall of M. tuberculosis, can also be recombinantly produced.^125,126 ESAT-6 can act both as a protective antigen against tuberculosis or can non-specifically enhance CMI to coadministered antigens.^52,125-127

CpG ADJUVANTS

Prokaryotic DNA contains unmethylated CpG dinucleotides within specific nucleic acid motifs that are recognized by the innate immune system of vertebrates.¹²⁸ These immunostimulatory motifs are the ligand for TLR-9 which is found primarily in intracellular vesicles of phagocytic cells.¹²⁸ Signaling through TLR-9 CpG ligands induces the production of reactive oxygen species and activation of NF-κB.¹²⁹ These immunostimulatory sequences are species specific and unique sequences have been described for laboratory animals (mice, rats, and rabbits), humans, and nonhuman primates, as well as companion and farm animals.⁶⁵ For humans, there have been two types of CpG motifs described, type K (also known as B-type) and type D (or A-type).¹²⁸ The type K CpG motifs primarily stimulate B cell and monocyte proliferation, IgM, IL-10, and IL-6 secretion. Type D CpG motifs primarily activate DCs, a response which is characterized by upregulation of CD80, CD86, MHC II, and TNF-α and IL-8 secretion.¹²⁸ Regardless, CpG motifs are capable of stimulating enhanced secretion of immunoglobulins, and may be capable of modulating preexisting immune responses.^64,129 Addition of CpG motifs to vaccine formulations has been shown to induce both cellular and humoral response to immunogens, inducing a Th1 bias. CpG motifs have been shown to induce demonstrable immune responses to weak immunogens such as malarial antigens, anti-H. influenzae glycoconjugates and melanoma antigens.¹²⁹ When both alum and CpG motifs were included in vaccine formulations, the resulting immune response was Th1-biased, with no IgE production or eosinphilia.⁸⁴ Furthermore, addition of CpG motifs to intranasal vaccine formulations enhanced the total serum titer to TT and influenza (viral) antigens in mice indicating that they may be useful as immune enhancers for mucosal delivery of antigens.¹³⁰ CpG motifs are also used to enhance the response to antigens encapsulated in biodegradable polymeric microspheres described in this review.^129,131,132 CpG motifs have been included in many experimental vaccines demonstrating enhanced protection against a variety of pathogens including Ebola virus, Bacillus anthracis, Francisella tularensis, L. monocytogenes, and Cryptococcus neoformans and in models of polymicrobial intraabdominal sepsis.^122,133-136

BACTERIAL TOXINS

Bacterial toxins have a high degree of immunogenicity and immune enhancing capabilities along with highly specific cellular receptors. These properties have led researchers to study the potential of bacterial toxins as vaccine adjuvants. Pertussigen, a complex mixture derived from B. pertussis, including pertussis toxin, has been used experimentally as an adjuvant.¹³⁷ Pertussigen enhances levels of IgE and hypersensitivity reactions to codelivered antigens and may help adjuvant the response to TT and diphtheria toxoid which are part of the trivalent childhood DPT vaccine.^6,137 Heat-labile enterotoxin (LT) from E. coli has also been shown to enhance mucosal immunity to coadministered antigens.³⁵ LT exhibits adjuvant efficacy for induction of mucosal and parenteral immunity in mice. LT was also used as an oral adjuvant for Campylobacter killed whole-cell vaccines. In rhesus monkeys, LT was shown to be safe and provided superior performance over the Campylobacter killed whole-cell vaccines alone.¹³⁸ Cholera enterotoxin (CT) is another bacterially derived protein that shows high immunogenic potential when delivered to mucosal surfaces.^139,140 LT is highly homologous to CT, but CT stimulates predominantly Th2 responses to conjugated antigens while LT stimulates mixed Th1/Th2 response.¹⁴¹ However, cholera-like toxin adjuvants delivered by the nasal route have been found to be taken up by the olfactory nerve and the central nervous system, leading to potential unwanted side effects¹⁴² and CT can induce diarrhea in humans. Not much is known about the cell-mediated immunity or delayed hypersensitivity response to CT. The ability of CT to act as a mucosal adjuvant has been confirmed by many investigators with a variety of antigens, and administering CT by a route different from the antigen is not immune enhancing.^139,140

CYTOKINES

The cytokine network controlling immunity and T cell development is complex and much research remains to be done to elucidate these pathways.¹⁴³ The effect of a few cytokines and their relevance to immune activation has been well studied and these cytokines have been explored as adjuvants to provide potentially less toxic approaches to enhancing vaccine efficacy. For example, granulocyte macrophage-colony stimulating factor (GM-CSF) has been included in experimental vaccines due to its ability to enhance APC recruitment and activation.¹¹⁶ In attempts to improve the pneumococcal polysaccharide vaccine against S. pneumoniae, IL-12 was included as a mucosal adjuvant.¹⁴⁴ The inclusion of IL-12 enhanced mucosal and systemic IgG2a and IgA following intranasal vaccination and showed a marked reduction in bacterial nasal carriage and prevention of bacterial systemic invasion.¹⁴⁴ Inflammatory cytokines in the IL-1 family have been shown to enhance the production of serum and mucosal IgG and IgA antibodies and cell-mediated responses to codelivered Ova and tetanus toxoid.³⁵ The choice of cytokine included in a vaccine formulation must be chosen with care. In a recent study, polylactide microspheres were investigated as intranasal delivery of recombinant V antigen (rV) of Y. pestis coencapsulated with IL-6, IFN-γ, or IL-4.¹⁴⁵ While all formulations induced mucosal IgG1 and IgA antibodies, only formulations including IL-6 with the rV induced protection from systemic bacterial challenge.¹⁴⁵ The challenge of cytokine delivery is the rapid utilization of cytokines and their pluripotent biological effects. One mechanism to reduce these effects is to deliver a plasmid including the sequence of the cytokine.^66,116 With DNA based vaccine technologies, this has proven very effective for enhancing the immune response to the DNA-encoded antigen. Inclusion of the sequence for IL-2 or IL-12 with the sequence with HIV antigen enhanced the production of a strong Th1 immune response.³⁵

POLYMER VACCINES

Biodegradable polymers have been studied for many years because they show promise for the development of single dose vaccines.^146,147 Polymers have the ability to sustain the release of the vaccine antigen over an extended period of time, thus eliminating the need of subsequent doses of vaccines. Other potential advantages of these materials are that immunomodulatory properties (i.e., adjuvanticity) can also be achieved with the proper tailoring of the polymer chemistry.¹⁴⁸ Studies evaluating the use of controlled-release, single dose polymeric vaccines in both laboratory animals and livestock species (i.e., sheep, mini-pigs, cattle, and horses) have shown promise when encapsulating protein antigens.^149-153

Biodegradable polymers also offer the advantage that MPLA, CpG DNA motifs or other immunomodulatory molecules can be incorporated to create a pathogen-mimicking solid particle.¹⁵⁴ Polymeric vaccine particles have been shown to induce demonstrable immune responses when administered by several routes including parenteral (e.g., intramuscularly or subcutaneously), intranasal, or oral.³⁵

These materials also have the added advantage over stable (nondegradable) devices (e.g., pumps) in that after administration, there is no need to remove them, therefore eliminating another surgical procedure. Furthermore, most are manufactured from synthetic parent compounds, eliminating many potential reactive antigenic or allergenic epitopes that can accompany the use of animal or plant derived materials.

The two most widely studied polymer classes for controlled release vaccines are polyesters^155-164 and polyanhydrides.^87,165-176 Other classes of polymeric compounds have been evaluated and shown to successfully deliver antigen to laboratory animals.^177-190 Key findings of research done with these polymeric systems as vaccines carriers are discussed below and the chemistries are shown in Table 4.

Table 4.

Structure of Biodegradable Polymers Studied for Use as Vaccine Adjuvants

Polymer	Structure	Refs.
Polysaccharides
Dextran		229
Chitosan		35,230
N-trimethyl chitosan		230
Polyanhydrides
Poly(sebacic acid) SA		148,214,219,226
1,3-bis(p-carboxyphenoxy)propane CPP		174,226
1,6-bis(p-carboxyphenoxy)hexane CPH		148,214,219
l,8-bis(p-carboxyphenoxy)-3,6-dioxaoctane CPTEG		217,221,222
Poly(trimellitylimido-l-tyrosine)		226
Poly(ortho ester)s		60
Poly(ester-amide)s
Phenylalanine-based PEA		235,236
Leucine-based PEA		235,236
Polyesters
Poly(lactic acid) LA		157,159,203
Poly(glycolic acid) GA		157,159,203
Poly-ε-caprolactone		225
Poloxamers
Poly(ethylene glycol) PEG		237,238
Other
Poly(vinyl methyl ether-alt-maleicanhydride) PVM/MA		224

POLYESTERS

Microspheres composed of polyesters have been the most widely studied. Polymers of lactic acid and glycolic acid (e.g., poly(lactide-co-glycolide), PLGA) have been utilized in biomedical applications such as bone pins and dissolvable sutures for many years and recently have proven effective as vaccine delivery vehicles for the induction of protective immunity in laboratory animals.^155-159 The greatest benefits of PLGA is that its degradation products, lactic acid, and glycolic acid, are naturally occurring metabolites and are readily absorbed by neighboring cells.^160,161 However, as the polyester degrades and the acidic monomers are released, an acidic microenvironment is created. Prolonged exposure to aqueous or acidic environments has been shown to be detrimental to the stability and immunogenicity of proteins, especially the proteins used in recombinant and subunit vaccines, for example, TT and DT.^162,163 Some attempts to minimize this acidity have been recently evaluated by incorporating a basic compound like magnesium carbonate (MgCO₃) into PLGA microspheres.¹⁶⁴ However, subsequent analysis indicated that while MgCO₃ did not significantly improve peptide stability, it did enhance the antibody production, acting as a potential adjuvant.

Antigen-loaded PLGA microspheres function as an adjuvant by at least two mechanisms: (1) creating a depot for the antigen in vivo, and (2) enhancing phagocytic uptake of the antigen-loaded particle by APCs.¹⁵⁹ The uptake of PLGA microspheres by macrophages or DCs has been demonstrated following administration by intraperitoneal or intradermal routes, respectively.¹⁹¹ Other immunostimulatory properties of PLGA were observed in studies showing an enhanced cytokine production and proliferation when cells were incubated in vitro with blank PLGA microspheres.¹⁶¹ Similarly, oral administration of PLGA nanoparticles containing type II collagen promoted the induction of tolerogenic immune responses that ameliorated arthritis.¹⁹² The prolonged presence of the nanoparticles in the Peyer’s patches and the induction of elevated TGF-β suggested the differential activation of DCs that modulated the subsequent immune response.

Vaccine formulations based on PLGA, PLA, or PGA variants have been successful in inducing immune responses in laboratory rodents to a large number of antigens including: Y. pestis antigens, HIV gp140, B. pertussis antigens, measles virus antigen, OVA antigen, TT, diphtheria toxin, type II collagen, malarial antigens, cancer cell antigens, E. coli adhesion proteins, Vibrio cholerae antigens, influenza virus antigens, hepatitis B viral antigens, and ricin toxoid.^{35,161,192,193} These vaccines have been delivered by a variety of routes including intradermally, intravaginally, intranasally, orally, or parenterally into laboratory animals to induce both serum antibodies, mucosal IgA, cell-mediated responses and facilitated the induction of secondary immune responses (e.g., isotype switching) as determined when individuals were analyzed up to a year after single immunization.^161,194 Many groups have reported the successful induction of immunity following use of a single dose vaccine formulation composed of PLGA microspheres of various compositions.^157,195-200 Furthermore, encapsulation of antigens in PLGA microspheres was shown to enhance antigen presentation via MHC I leading to increased activation of antigen specific cytotoxic T cells.^147,193,195 However, most of these studies were conducted in vitro, and some investigations included MPLA, a known Th1 immune response activator, in the microsphere while others used multiple injection regimens in vivo. There is no consensus opinion, however, as to whether PLGA-based vaccines are more efficacious than current adjuvant systems such as alum. Antibody responses induced in mice and guinea pigs following vaccination with TT-loaded PLGA were greater than those induced by single injection of soluble TT alone or two doses of alum absorbed TT. Additionally, a stronger anamnestic response (higher titer) was observed when individuals that had received the TT-loaded PLGA microparticles were boosted 1 year later.¹⁹⁴ On the other hand, Walker et al.¹⁹⁹ observed that encapsulation of TT in PLGA microspheres did not induce serum antibody titers higher than alum-based TT vaccines. Only small amounts of antigenically active TT were released in the first 2 days from PLGA microspheres, even though protein continued to be released for up to 11 weeks.¹⁹⁴ Collectively, evaluation of PLGA studies does not provide strong correlation between release of antigenic peptides, length of in vitro release of peptides, and immune response to those peptides in vivo.

Some studies have suggested that immunization with PLGA microspheres effects immune deviation. Moore et al.²⁰¹ showed the ability of HIV gp120 protein loaded PLGA microspheres to shift the T cell response from a dominant Th2 or mixed Th1/Th2 to a more dominant Th1 immune response as indicated by the presence of IFN-γ producing CD4⁺ T cells. In other studies, the Th2-biased hepatitis B core antigen has been formulated with the Th1 immune stimulator MPLA in PLGA nanoparticles to develop a stronger Th1 response.⁶⁸ More recently, a vaccine formulation prepared against malaria and composed of PLGA microspheres and Montanide ISA-720 was shown to induce an antibody response (IgG isotype class switching) characteristic of Th1 response.²⁰²

Variations in reported efficacy of PLGA microspheres may be due to dose of antigen, method of encapsulation (e.g., spray drying vs. solvent evaporation), route of immunization, and/or the size of the microspheres.^193,203 Following primary immunization with small microspheres (10–20 μm), a greater anamnestic response was generated 1 year later following a low dose booster than that observed in animals initially receiving larger microspheres (>60 μm);¹⁹⁴ however, nanoparticles (200–600 nm) were less effective at inducing cell-mediated immune response than microspheres.¹⁹³ This may be because microspheres <10 μm in diameter are readily phagocytosed by macrophages and DCs that would enhance antigen processing and presentation.^204-209 On the other hand, the route of immunization with PLGA microparticles influenced the type of immune response generated. The intraperitoneal route induced Th1 cell-mediated response while the intramuscular route induced a Th2 humoral response.¹⁹³ Despite all the extensive research done with PLGA as antigen carriers, some with success in animal models, no formulation has been reported to induce a protective immunity in humans.²¹⁰

POLYANHYDRIDES

Polyanhydrides are a class of surface erodible, biocompatible polymers that have been extensively used as carriers for controlled drug delivery.^87,165-176 These biodegradable polymers are currently approved by the FDA for use in a variety of biomedical applications and can also be fabricated into protein-loaded microspheres.²¹¹ Biocompatibility studies have shown that these biomaterials degrade into carboxylic acids, which are nonmutagenic and noncytotoxic products.^212,213 The surface erosion mechanism leads to a controlled release profile with predictable degradation profiles, which can range from days to months, depending on the copolymer composition.^214,215 In addition, studies involving polyanhydride delivery systems for vaccines have shown attractive features such as improved adjuvanticity, antigen stabilization, and enhanced immune responses.^{165,175,176,216}

The main advantage of polyanhydrides over polyesters as antigen carriers is associated with the enhanced protein stability following encapsulation. Studies have shown that polyanhydrides are capable of stabilizing polypeptides and sustaining their release without the inclusion of potentially reactive excipients or stabilizers.^217-220 The hydrophobicity and surface erosion characteristics of polyanhydrides prevent water from penetrating to the interior of the microsphere thus preserving the encapsulated antigen in its native state (i.e., increased stability). Furthermore, the degradation products of polyanhydrides are less acidic than those of polyesters, which may further enhance the stability of encapsulated antigens and reduce tissue reactions to the polymer.^217,219 Despite these beneficial characteristics, the use of polyanhydrides for vaccine delivery has not been extensively evaluated.

Recently, Kipper et al.¹⁴⁸ performed in vivo studies to evaluate the induction of immune responses following immunization with antigen-loaded microspheres based on the anhydride monomers, sebacic acid (SA) and 1,6-bis(p-carboxyphenoxy)hexane (CPH). Microspheres encapsulating TT antigen were injected in C3H/HeOuJ mice. These studies demonstrated that TT maintained its immunogenicity and antigenicity following encapsulation. The type of immune response generated, Th1 versus Th2, was evaluated by antibody isotypes. It was observed that TT loaded 20:80 CPH/SA microspheres enhanced the immune response after a single dose and induced a dominant Th2-like immune response. However, the 50:50 CPH/SA produced a balanced Th1/Th2 response. Total TT-specific IgG titer remained high regardless of dominant isotype. The preferential enhancement of the Th1 immune response resulting in more balanced immune response (i.e., immune deviation) is a unique and valuable feature of this delivery vehicle that makes it a promising adjuvant candidate for vaccines. Currently, the groups led by Narasimhan and Wannemuehler are corroborating the immunomodulatory properties of the CPH/SA system with other antigens as well as investigating the adjuvant properties of novel amphiphilic polyanhydride chemistries. Copolymers of CPH and 1,8-bis(p-carboxyphenoxy)-3,6-dioxaoctane (CPTEG), which contains ethylene glycol moieties in the polymer backbone, are promising candidates for the development of vaccines as they have been shown to provide a conducive environment for protein stabilization.^217,221-223

Anhydride monomers have been copolymerized with other chemistries and their potential as adjuvants have been evaluated. An immunogenic subcellular extract obtained from heat-killed Salmonella enteritidis cells (HE) has been encapsulated in nanoparticles of a copolymer comprised of methyl vinyl ether and maleic anhydride (PVM/MA), best known as Gantrez^® polymer.²²⁴ In this study, 80% of the Gantrez^®-HE immunized mice survived even when the nanoparticle formulation was administered 49 days prior to a lethal challenge. As early as 10 days after immunization, a Th1 immune response was demonstrable in these mice as determined by the IgG2a antibody titer in the serum. On the other hand, a dominant Th2 immune response was present at 49 days after immunization (IgG1 > IgG2a). Since it is known that a Th1/Th2 balance is required to protect against S. enteritidis infection, this copolymer is a promising candidate for the development of future vaccines. In this regard, blank nanoparticles of Gantrez^® administered subcutaneously four weeks prior to challenge induced a level of protection similar to that induced by antigen-loaded nanoparticles or the Rv6 commercially available vaccine against S. enteritidis serovar abortusuis.²²⁵ While the authors did not demonstrate the presence of antigen-specific immunity, these data suggest that the blank nanoparticles were able to induce and sustain sufficient innate immunity to provide nonspecific protection against subsequent Salmonella infection. In the same study, abortusovis antigen-loaded poly(ε-caprolactone) microparticles did not induce protection.

In another attempt to design suitable carriers specifically intended for vaccine delivery, Hanes et al.²²⁶ synthesized poly(anhydrides-co-imides) with the adjuvant l-tyrosine incorporated in the polymer backbone. In these studies, a predictable and controlled protein release was observed from microspheres of poly[trimellitylimido-l-tyrosine-co-sebacic acid-co-1,3-bis(p-carboxyphenoxy)propane] and polymeric implants were well tolerated after subcutaneous implantation in rats. More recent studies demonstrating the suitability of polyanhydrides for use in single dose vaccines involved the design of a core-shelled cylindrical device composed of a biodegradable hydrophobic coating and laminated core of polyanhydrides and polyphosphazenes.²²⁷ Polyanhydrides based on SA were used as isolating layers of the cylinder in order to produce a pulsatile drug release, a mechanism which would minimize doses of vaccines. Even though these polyanhydride systems showed promising characteristics for vaccines design, no further in vivo studies evaluating the characteristics of the proposed adjuvant were validated.

A comparative study between polyanhydrides and polyesters has demonstrated the potential capabilities of polyanhydrides for oral vaccination.²²⁸ Microspheres (0.1–10 μm) composed of fumaric acid (FA) and SA proved to have strong adhesive interactions with the mucosal gastrointestinal lining of rats, as opposed to poly(lactic acid) (LA), which showed minimal uptake. The adhesive interactions are ideal to prolong the biological activity of the delivered antigen or bioavailability of encapsulated drugs. Not surprisingly, plasmid DNA- and anti-coagulant drug dicumarol-encapsulated FA/SA microspheres enhanced gene activity and plasma drug levels, respectively, when compared to the controls. In the same studies, blends of FA and LA were used for insulin delivery and groups that received the formulation were able to regulate glucose levels as opposed to the groups that received insulin only. Even though the biological activity of insulin was preserved, it was the adhesive characteristic of FA the responsible for the efficient delivery.

OTHER POLYMERS

Naturally Derived

Several naturally derived polymeric materials, such as dextran, chitosan, starch, and alginate have been evaluated in laboratory models for use as vaccine adjuvants. In the case of dextran, it has been chemically modified or use in conjunction with other adjuvants in order to improve its immunogenicity. Immunization of cattle with dextran in combination with mineral oil against Streptococcus bovis and Lactobacillus spp. induced the highest serum IgG responses when compared with other adjuvants (i.e., FCA, Quil A, alum), presumably due to the combined effect of both substances.¹⁷⁷ In studies involving vaccination of cattle against M. tuberculosis, diethylaminoethyl (DEAE)-dextran induced high levels of IL-2 and low levels of IFN-γ, indicating a strong humoral response not desirable for this particular disease.¹⁷⁸ Interesting results were obtained when a dietary supplementation of Lactobacillus casei with dextran enhances humoral immune responses, and chickens were able to maintain the growth of the bacteria in their intestines and prevent possible infections.¹⁷⁹ Vaccines that have been evaluated utilizing cross-linked dextran microparticles, containing conjugated TT induced serum antibody to TT for long periods, eliminating the need of additional booster doses.²²⁹

Chitosan, a cationic polysaccharide derived from chitin in the exoskeleton of crustaceans, can also be formulated into microparticles capable of encapsulating antigen.²³⁰ Studies with chitosan showed that the immune bias induced by vaccination with antigen containing chitosan microparticles was more dependent on the route of delivery (e.g., intranasal vs. parenteral) than the nature of this adjuvant.^35,180 An intranasal delivery of N-trimethyl chitosan chloride (TMC) containing diphtheria toxoid enhanced the immune response when compared with the conventional alum adsorbed vaccine.²³⁰ This enhancement of nasal vaccination is likely a result of the mucoadhesive properties of chitosan, which enhances penetration across nasal mucosae.^181,182 More recent studies with chitosan and TMC establish that chemical variables, such as molecular weight in chitosan and degree of quaternization in TMC influence the magnitude of the immune response after nasal administration.¹⁸³

Another natural polymer with potential in vaccines is starch, which also has been assessed in mucosal vaccines. Some advantages of starch include its inert properties, proven safety, and commercial availability.¹⁸⁴ Heritage et al.¹⁸⁵ found that human serum albumin delivered on starch microparticles grafted with polydimethyl-siloxane stimulated systemic and mucosal immune responses. Similarly to studies done with chitosan, the route of administration of starch influences the immune response.¹⁸⁶ Among oral, subcutaneous, and intramuscular administrations, vaccines delivered by the subcutaneous routes induced stronger humoral responses. However, when comparing oral and intramuscular routes, stronger humoral response was induced after oral primary administration and a stronger cell-mediated response after oral booster doses. Although the adjuvant capabilities of starch were proved with success in mice studies, a human vaccine trial was not successful.²³¹

Alginate microparticles offer several advantages for vaccine applications, including good biocompatibility, ease of preparation, and antigen protection during fabrication and administration.^184,232 Alginate microparticles have been administered to several animal species (i.e., mice, rabbits, cattle, and chicken).²³² The enhancement of the immune response induced in the animals after oral administration with antigen-loaded alginate microparticles shows promise for the development of veterinary vaccines. Nevertheless, in vitro studies show that alginate is not the optimum chemistry to activate human-derived DCs, as it decreases the expression of costimulatory molecules and antigen presenting complexes when compared to nontreated cells.²³³ Other in vitro studies that simulated gastric fluid environment showed that alginate microparticles were not able to stabilize live rotavirus vaccines.²³⁴

Synthetic

Other novel polymer chemistries have been researched to overcome the limitations of available polymers as vaccine carriers. The novel poly(ester-amide) (PEA) copolymers, composed of amino acid residues, diols, and dicarboxylic acids, have been shown to enhance cellular immunity.¹⁸⁷ Polyamide gives PEA its superior mechanical and thermal properties, while the polyester portion is responsible for its flexibility and hydrolytic susceptibility, allowing PEA to degrade within a reasonable period of time. It is biodegradable, however, in contrast to polyester and polyanhydrides, PEA degrades by enzymatic cleavage within the body.^235,236 Thus, shelf life and handling does not affect its degradation rate and the polymer remains intact until needed for therapy. PEA has been conjugated with several therapeutics peptides, including human melanoma antigen-derived peptides (MART), a synthetic peptide based on the gp120 protein of HIV, and a MHC II-restricted T-cell epitope from the influenza A virus hemaglutinin (HA) protein.¹⁸⁷ In general, the studies evaluating PEA-peptide conjugates demonstrated that cellular immunity, encompassing both MHC I- and MHC II-restricted T-cell responses, was enhanced.

More recently, in vivo studies in mice have shown that poly(ethylene glycol)-stabilized poly (propylene sulfide) nanoparticles target the APCs directly in the lymph nodes.^237,238 In these studies it was found that particles in the size range of 20–45 nm enter lymphatic vessels and subsequently target DCs in the lymph nodes. The cross-linked polymer system used here degrades into a water soluble polymer under oxidative conditions.

Polymers in Plasmid DNA Vaccines

Plasmid DNA vaccines represent a promising alternative against intracellular pathogens. Even though plasmid DNA immunogens have elicited strong cell-mediated responses in small laboratory animals, these have not had success in limited human clinical trials.¹⁸⁸ Ideal adjuvants will improve the magnitude of plasmid DNA expression, must protect DNA from enzymatic degradation, and must facilitate the DNA plasmid uptake into cells. Several polymer chemistries have been evaluated in conjunction with DNA vaccines and a thorough discussion of this topic is beyond the scope of this review. In short, microspheres of polyesters, polycarbonates, polystyrene, and poly(orthoesters) have been used in DNA vaccination and their administration resulted in enhanced immune responses when compared to naked DNA administrations.^60,189,190 Table 5 summarizes the pros and cons of polymeric adjuvants.

Table 5.

Advantages and Disadvantages of Polymers as Vaccine Adjuvants

Polymer	Antigens	Advantages	Disadvantages	Refs.
Polysaccharides
Dextran	Streptococcus bovis Lactobacillus spp Mycobacterium tuberculosis Tetanus toxoid (TT)	Induces strong humoral responses	Not desirable for some diseases (e.g., tuberculosis)	101
Chitosan N-trimethyl chitosan	Diphtheria toxoid	Enhanced immune response compared to alum Mucoadhesive properties		230 35
Polyanhydrides	TT Salmonella enteritidis (HE) Plasmid DNA	Biocompatible nonmutagenic, noncytotoxic degradation products Degradation products have low acidity Enhanced protein stability Immunomodulatory Mucoadhesive properties	Processing and storage	90,113,214,219,217,221,222,226
Poly(orthoester)s	Plasmid DNA	Enhanced immune response when compared to naked DNA	Plasmid DNA unsuccessful in human clinical trials	60
Poly (ester-amide)s	Melanoma antigen derived peptides (MART) HIV gp120 MHCII restricted T-cell epitope from influenza A virus hemaglutinin (HA)	Degrades by enzymatic cleavage Enhanced cell-mediated) immunity		235
Polyesters	TT Diphteria toxoid Yersinia pestis HIVgp140 Bordetella pertussis Measles virus antigen Ovalbumin Type II collagen Malarial antigens Cancer cell antigens Eschericia coli Ricin toxoid Vibrio cholerae Influenza virus antigens Hepatitis B viral antigens Plasmid DNA	Degradation products are biocompatible and easily metabolizable Antigen-loaded microspheres enhance uptake by APCs Experiments have shown increase in both humoral and cellular immune responses	Acidic microenvironments detrimental to antigens No protective immunity in humans have been reported yet Poor mucoadhesive properties	57,99,101,159,161
Poly(ethylene glycol)		Can target APCs in LNs		237,238
Poly(vinyl methyl ether-alt-maleic anhydride)	Salmonella enteritidis (HE)	Th1/Th2 balance Nonspecific protection against Salmonella		224

THE IDEAL VACCINE ADJUVANT

Vaccines and their adjuvants interact with the patient’s immune system in a variety of ways. Thus, there is no single set of characteristics that would describe an ideal vaccine adjuvant for all situations. An adjuvant must be appropriate to the particular delivery route (e.g., intramuscular, mucosal, intraperitoneal, etc.), desired immune response (cell-mediated vs. humoral), pathogen, and stage of a disease. Additionally, biological traits of the patient may also be important including species, race, age, medical history, and genetic makeup. All of these factors may influence the effectiveness of a vaccine adjuvant, and the effects of these factors may be unknown. Nonetheless, there are certain characteristics that a good vaccine adjuvant must possess. These characteristics can be broadly grouped into two categories: biological characteristics and practical or economical characteristics.

Because vaccine adjuvants may enhance the immune response through different modes of action, the particular mechanism of adjuvanticity is of paramount importance. The mechanisms of adjuvant activity have been classified in different ways by different authors.^66,119,239 The broadest classification distinguishes among two types of mechanisms: immune stimulation and targeting antigens to particular cells or tissue types.⁶⁵ Adjuvants which act through the later mechanism target vaccines to DCs, through interactions with transmembrane TLR proteins or other cell surface receptors,⁹⁷ or by virtue of their size.^240,241 Polymer microspheres and liposomes <10 μm in diameter may be readily phagocytosed by macrophages and DCs.⁹⁷ This specific targeting can reduce the quantity of antigen required to induce protective immunity. A good immunostimulatory vaccine adjuvant must stimulate the desired immune response without toxicity or inducing excessive inflammation. While some immunostimulatory adjuvants of bacterial origin have potent adjuvanticity (e.g., LPS), they can also be extremely toxic (e.g., induction of tumor necrosis factor).²⁴² Less toxic adjuvants, such as alum, may also be less potent or ineffective at eliciting cell mediated immunity.^97,243,244

Good immunostimulatory vaccine adjuvants activate DCs to mature into APC and migrate to the draining lymph node, coincident with induction of the cytokine profile appropriate to the desired immune response mechanism (i.e., IFN-γ, IL-2, and IL-12 for the Th1 response and IL-4, IL-5, and IL-6 for the Th2 response). Like adjuvants that target DCs, some immunostimulatory vaccine adjuvants also interact with TLR proteins. Though these proteins have affinity for a variety of ligands, different subpopulations of DCs express different TLR profiles and, thus, have different degrees of sensitivity to different antigens and adjuvants.^31,245,246 Furthermore, the same TLR may activate different intracellular signaling cascades leading to different activated phenotypes in different DC subpopulations. Regardless of the mechanism of adjuvanticity, vaccine adjuvants must activate this desired adaptive immune response without over stimulating innate immune function.

Economical and practical considerations must also be taken into account when selecting an ideal vaccine adjuvant. Singh and O’Hagan^64,74 list biodegradability, ease of manufacture, and low cost among important characteristics of vaccines. Other practical aspects to be considered include stability over time, ability to provide immunity with a single dose, and suitability for mucosal delivery. Such characteristics would enable more practical and economical strategies to fight infectious disease in remote areas that lack developed public health infrastructure and in communities that do not have access to modern medical care.²⁴⁷ Finally, while the “depot” effect (long thought to be the primary mechanism of adjuvanticity for alum) is no longer regarded as the essential mechanism behind adjuvant effectiveness,^97,248 formulations such as degradable polymer microspheres may provide sustained exposure to antigens, obviating the need for multiple administrations. Practical considerations such as stability and cost may preclude the widespread use of some otherwise potent protein adjuvants such as cytokines.^64,244

Thus, we can summarize the ideal vaccine adjuvant as one which selectively targets the antigen to the desired population of APCs, minimizes the amount of antigen required, induces the desired adaptive immune response while minimizing the innate immune response, is minimally toxic, low-cost, stable for long-term storage, and provides protective immunity in a single dose via a convenient delivery route.

NEW RESEARCH TOOLS TO STUDY DISEASE PREVENTION

New adjuvants are also needed that can be used to precisely tune the nature or outcome of the immune response to more effectively protect against particular diseases such as cancers and HIV. This may be done by controlling the induction of particular cytokine profiles and by more effectively targeting antigens to specific tissues, cells, or intracellular compartments (e.g., DNA vaccines to the nucleus of a cell). These new adjuvants could also be used as research tools to study the induction or regulation of different immune response mechanisms that are associated with autoimmune diseases, allergies, or tolerance. Excitedly so, many of these new adjuvants are being developed experimentally and much more research is needed to bring them to an application. As shown in Table 1, very few adjuvants are being used in licensed vaccines. Even the materials being tested in current clinical trials represent relatively few new immunostimulating adjuvants or chemistries, especially against infectious diseases (Tab. 2). Furthermore, there may be a need for a considerable shift in thinking about how vaccines are tested for efficacy. Antibody titer is almost universally used as the test for vaccine efficacy but often high antibody titers do not translate into the best protection.²⁴⁹ Many times a highly immunogenic antigen does not correlate to a protective immune response. This “deceptive imprinting” is a common evasion mechanism by pathogens and partially responsible for the slow development of HIV vaccines.²⁵⁰ Also there is the caveat that laboratory mice are not humans (or other livestock species) and what works in a mouse may not translate to other species. Numerous studies have highlighted differences in mouse and human immune systems including differences in complement reactivity,²⁵¹ induction of Th17 cells,²⁵² or response to a vaccine based on particle size.⁷⁰

In 2006, the National Research Council convened a Workshop on Immunomodulation²⁵³ which made several recommendations to improve vaccine design including (1) an improved molecular level understanding of the innate immune system, (2) the need for effective delivery mechanisms, (3) the identification of potential molecular targets to modulate innate immunity without undesirable side effects, and (4) new strategies to target DCs and optimize antigen presentation. A key need that was identified by this panel was that in order to solve these important problems, it is critical for researchers from multiple disciplines to work together. These fields may include biochemistry, immunology, materials science, cell biology, computational biology/materials science, pathology, oncology, microbiology, and combinatorial science. It is important to combine expertise from antigen biochemistry, cell biology, and immunology to understand the mechanism of immunogenicity and how the preservation of various epitopes contributes to immunogenicity. As these antigens are combined with adjuvants, it is important for materials scientists to work closely with immunologists to understand how protein antigens can be stabilized during encapsulation and delivery and how adjuvants interact with APCs. As these adjuvanted systems enter the body, they encounter plasma proteins that may adsorb on to the surface of the adjuvant. How this affects the release of the antigen and how this influences APC activation or antigen processing is of great significance to the initiation of the desired immune response. Finally, the use of the appropriate animal models to study these phenomena is critical and immunohistochemical methods are needed to study how these adjuvants affect the local tissue response.

In this regard, the authors, who belong to chemical engineering and veterinary microbiology departments have worked towards providing a highly cross-disciplinary research environment for students and postdoctoral researchers in their respective groups. The chemical engineering graduate students have the opportunity to take courses on immunology and molecular biology techniques, participate in journal clubs, and several of them have completed an immunobiology certificate program on their way to a Ph.D. Likewise, the microbiology students have the opportunity to take courses on polymeric biomaterials and nanotechnology. Such an approach has immensely benefited students from both disciplinary groups and has prepared them to address diverse research problems with new and innovative perspectives. Similar examples of cross-disciplinary research groups exist and are much needed as scientists embark on new therapies for diseases such as cancer, HIV, and respiratory infections.

Over the last 200 years, the use of vaccines has proven to be one of the most successful medical interventions in the reduction of disease caused by infectious agents.¹ However, many challenges still remain with regard to fully realizing the health benefits of active immunization programs. Some of these obstacles include the implementation of improved adjuvants, development of single dose vaccines, methods to overcome the poor immunogenicity of recombinant and subunit immunogens, and the ability to rapidly and rationally develop vaccines against emerging pathogens. In this regard, the mechanisms underpinning the effective modulation of cellular and molecular events associated with adjuvant enhancement of immune responses are still largely unknown. There is growing interest in the development of vaccine delivery systems based on micro- and nano-scale devices composed of biodegradable polymers, because they have the potential to act as effective adjuvants by encompassing all three of the classical adjuvant properties: providing an antigenic depot with a tailored and pulsatile release of the antigen over time, directing particulate antigens to the APCs and modulating the activation of innate immunity by altering polymer chemistry.⁹⁵ However, the mechanism of adjuvanticity and the ability of adjuvant chemistry to selectively modulate the immune response are still largely unknown. In order to address these challenges, it is important to perform fundamental and systematic studies of the role of polymer chemistry in regulating activation of APCs (e.g., DCs), antigen uptake, processing, and presentation, migration to the draining lymph node, and modulation of the immune response.

The mechanisms by which adjuvants enhance and/or redirect the immune response (e.g., formation of high titer antibodies, CD4⁺ helper T lymphocytes and/or CD8⁺ T lymphocytes) in order to establish long-term immunologic memory are poorly understood. Upon antigen stimulation, T cells differentiate into two distinct populations described as Th1 and Th2 type immune responses.²⁵⁴ Furthermore, Th1- and Th2-related cytokines (IFNγ or IL-4/IL-13 respectively) can impact both the quality and magnitude of humoral and cell-mediated immunity. Humoral immunity, characterized by the activation of B cells that differentiate into antibody secreting plasma cells, is effective at neutralizing toxins, viruses, complement fixation, and opsonization of extra-cellular pathogens whereas the cell-mediated immunity (i.e., activation of cytotoxic T cells and macrophages) are crucial for protection against intracellular pathogens.²⁵⁵ The balance of humoral and cell-mediated immune responses has been shown to be important in the favorable outcome of many disease states. In this regard, vigorous and inappropriate cell-mediated immune responses have been implicated in the induction of autoimmune and chronically inflammatory diseases (multiple sclerosis and Crohn’s disease) while robust humoral immune responses are associated with allergic reactions.¹²⁰ In order to control the induction of appropriate immune responses and reduce the risk of autoimmunity or allergic responses, there is an urgent need to develop new, well-characterized adjuvants that allow for tailored immune activation and deviation. In spite of these implications of immune deviation, the mechanisms by which adjuvants influence whether Th1 or Th2 cells dominate an immune response are not well understood. Additionally, it is also important to consider the use of adjuvants to induce regulatory T cell responses and to avoid the aberrant induction of Th17 cells that have been associated with chronic inflammatory diseases.

Both in vitro and in vivo studies with the adjuvants discussed above indicate that adjuvant chemistry and particle size may play an important role in regulating the cellular and molecular mechanisms responsible for modulating host immune responses. Additionally, in order to understand intracellular trafficking at a molecular level, it is important to employ the use of reporter molecules (e.g., quantum dots (QDs)) embedded within the adjuvant. Encapsulating QDs within adjuvants will provide adequate intracellular stability to effectively track the transport of nanospheres through intracellular compartments. The luminescence properties of QDs are expected to persist so long as the integrity of their nanocrystal structure is maintained, providing superior performance compared to conventional fluorescent dyes such as FITC, whose fluorescence is sensitive to pH.^256,257 QDs can also dramatically enhance in vivo imaging of APC migration by using red and near-infrared emitting QDs as an alternative to Cy5, Cy5.5, or other traditional organic dyes. QDs have substantially larger absorption cross-sections than even the best commercial dyes developed specifically for such imaging applications.²⁵⁸ This improves the effective brightness of the fluorescence emission signal considerably. Additionally, QDs have unrivaled photostability that allows continuous long-term excitation without a substantial loss in fluorescence.²⁵⁶

Finally, in order to understand the cellular and molecular mechanisms that establish immunologic memory, it is very important to correctly choose appropriate in vitro/in vivo models that will promote the induction of cell-mediated as well as humoral (i.e., antibody) immune responses. Because there are likely to be subtle immunogenetic differences between mouse strains (and eventually individual human subjects), the use of combinatorial approaches evaluating cell–adjuvant interactions may provide a robust and versatile approach to the development of vaccines that will effectively stimulate immunity for different conditions and/or applications. These approaches may be used to rapidly screen a large number of adjuvant chemistries for their ability to differentially activate APCs, which will aid in the rational use of cocktails of micro- or nanoparticles in vaccine formulations. These formulations will possess the ability to stimulate the appropriate immune response depending upon the disease. The availability of transgenic models (e.g., OTI and OTII transgenic mice) provides for the capability to critically evaluate the activation of CD4⁺ and CD8⁺ pathways while other molecular biology tools enable researchers to evaluate the effect of new adjuvants on antigen processing and presentation both in vitro and in vivo.

In summary, an integrated and cross-disciplinary approach is needed that combines the development of novel adjuvants with: (i) molecular level studies that will elucidate the mechanisms of chemistry-mediated cellular activation by adjuvants; (ii) cellular level studies that will elucidate the uptake mechanisms of antigen-loaded adjuvants by immune cells and the activation and migration of these cells; and (iii) in vivo studies that highlight the underlying mechanisms governing immune modulation. Such an integrated approach is essential to solve the important challenge of rationally designing vaccine delivery systems that will effectively stimulate the immune system. It can provide new insights into the mechanisms of adjuvanticity and on the complex relationships between adjuvant chemistry, molecular mechanisms of APC activation, antigen uptake, processing/presentation by APCs, migration to the draining lymph node, and modulation of the immune response. To carry out such an approach, it is important to assemble highly cross-disciplinary teams of researchers with expertise in the areas of molecular and cellular immunology, intracellular trafficking, biomaterials chemistry, toxicology, nanotechnology, and pathology.