<< All Back-issues
<< This Issue's Table of Contents
ILAR Journal V37(3) 1995
Adjuvants and Antibody Production
| W. Carey Hanly, Ph.D., is an Associate Professor, Department of Microbiology and Immunology; James E. Artwohl, D.V.M., is a Clinical Veterinarian, Biologic Resources Laboratory; and B. Taylor Bennett, D.V.M., Ph.D., is an Assistant Professor and Director of the Biologic Resources Laboratory, University of Illinois at Chicago. |
In recent years the production of antibodies (Abs) in laboratory animals has become an essential part of many research projects. Investigators preparing to produce antibodies are confronted with a number of complex choices, some of which may be critical for success. The goal is to obtain high-titer, high-affinity antisera in a manner consistent with the welfare of the animals being immunized. Investigators not familiar with antibody (Ab) production require some guidelines to carry out immunizations in an appropriate manner. Recent studies of components of the immune system, their functions, and the actions of adjuvants on the immune system have yielded valuable information for understanding immune responses. The following discussion is an attempt to provide the non-immunologist with a rational basis for making decisions about choices of animal, immunization procedures, and adjuvants. An overview of the humoral immune response is included to provide a basis for understanding the function of adjuvants and immunization strategies. Additional information may be obtained from any of several immunology textbooks (Janeway and Travers, 1994; Paul, 1994). Although optimization of procedures is still largely empirical, successful production of useful Abs can usually be accomplished by following currently available guidelines.
THE IMMUNE SYSTEMS OF POULTRY AND MAMMALS
The immune system of vertebrate species can be broadly divided into the innate system and the adaptive (acquired) system. Interactions between the two are important for the function of both. The innate system, which appears to be evolutionarily older, lacks a high level of specificity and efficiency but responds rapidly. It consists of (1) cells, such as macrophages and dendritic cells, and (2) humoral factors, such as complement. The adaptive immune system has a high level of specificity and efficiency but it responds slowly upon first encounter with a given foreign antigen (Ag). However, it often establishes long-term memory. The adaptive system can be divided into (1) the cellular immune system and (2) the humoral immune system. Adaptive cellular immunity is based in T cells (thymus-derived lymphocytes) and functions primarily to identify and dispose of infected or mutated host cells. Humoral immunity resides in the Abs present in the fluid compartments of the organism. Antibodies, made and secreted by B cells, identify extracellular foreign material within the host and help to neutralize and dispose of it.
In mammals, Abs belong to one of five immunoglobulin (Ig) classes: IgM, IgG, IgA, IgD, or IgE. In avian species, they belong to one of three classes: IgM, IgY (also called IgG), or IgA (Warr et al., 1995). IgG and IgA may be further subdivided into subclasses in some species. All Ig molecules have a similar basic 4-chain subunit structure consisting of two identical heavy chains and two identical light chains arranged in H-L pairs (Figure 1). Each chain has a variable region and a constant region. VH- and VL-domains pair to form two identical Ag-binding sites per subunit structure, whereas domains of the heavy chain C-regions pair to form a segment of the molecule called Fc that imparts biological function. The V-regions differ in structure from one Ab specificity to the next, whereas the C-regions are relatively constant in structure within a given grouping of H-chains or L-chains. H-chain constant regions will differ between classes and subclasses (also called isotypes), but are identical between Abs of the same class or subclass within a species, with the exception of minor genetic variation. It will of course be understood that species-associated structural variations of Ig molecules occur, as with other proteins.
Ab molecules are bifunctional: V-regions determine Ag-binding specificity and C-regions determine biological effector functions (such as complement activation by the classical pathway and placental transport). The bifunctionality reflects the fact that the Ig H- and L-chains are each encoded by two genes, one for the V-region and one for the C-region. C-region genes are embedded in the germline genome and are limited in number. Functional V-region genes do not exist in the germline genome; instead they are assembled from banks of genomic DNA segments during the development of cells destined to become Ab-forming cells. The "mix-and-match" assembly results in great diversity of V-region genes. Only one functional VH[ gene and one functional VL gene is assembled in a B cell such that any given B cell specializes in making Abs of a single specificity. The phenomena of V-gene assembly and V-C gene pairing results in production of populations of Ab molecules with great diversity in Ag specificity and a limited number of effector functions.
THE HUMORAL IMMUNE RESPONSE TO PROTEIN AGS
The classic illustration of the serum Ab response of an animal to a soluble protein Ag shows two major phases: (1) a low-level primary response after the animal's first encounter with a foreign protein Ag, and (2) a much higher level secondary response after the animal's second encounter with the same protein. The primary Ab response has a lag period of 4 to 5 days after host Ag exposure, is predominantly of the IgM class, and is of short duration, peaking at 10 to 14 days and declining thereafter as the Ag is cleared. The secondary Ab response has a shorter lag period (3 to 4 days), is predominantly IgG, and not only has a higher peak titer, but also has a slower decline. The secondary response reflects, in part, the phenomenon of immunologic memory. When the primary response (IgM) changes to a secondary response with expression of IgG Ab molecules, a CK gene replaces the CT gene that encoded the IgM H-chain C-region in the primary response. Because the same H-chain V-region gene is used, the Ab specificity stays the same while the biological function changes.
The described response is to a soluble protein Ag administered without adjuvant. Soluble protein Ags administered by themselves are generally poor immunogens. However, the use of an adjuvant with the protein immunogen will increase the serum Ab titer and will also cause a more prolonged response as illustrated by the studies of Herbert (1968) (Figure 2). Responses to non-protein Ags are, in most cases, essentially primary responses irrespective of the number of times Ag is administered to an animal; that is, there is no memory, although persistence of Ag may cause a prolonged response. Moreover, adjuvants are much less effective with non-protein Ags. For this reason the focus of this review is on protein or peptide Ags as these Ags can induce high-titer, high-affinity serum Ab responses of long duration.
THE CELLULAR BASIS OF THE HUMORAL IMMUNE RESPONSE
The serum Ab response reflects activation and regulatory events that occur at a cellular level. Antibodies are synthesized and secreted by short-lived plasma cells, which are terminally differentiated B-lineage cells (bone marrow-derived lymphocytes in mammals and bursa-derived lymphocytes in poultry). Plasma cells derive from B cells that have been activated by Ag through their membrane-anchored Ag receptors (surface bound Ab molecules). For protein or peptide Ags, the activation of B cells is complex and requires participation of several other cell types, including T-helper cells (TH), dendritic cells, and macrophages. Thus protein Ags are called T-dependent (TD) Ags. Non-protein Ags are called T-independent (TI) Ags because participation of T-cells is not required. It should be noted that TI Ags can be converted to TD Ags--with the accompanying advantages of being able to induce high-titer, high-affinity serum Ab responses with memory--by conjugating them to protein carriers.
The Clonal Nature of Lymphocytes
B cells and T cells are responsible for specificity in the adaptive immune system. These cells recognize Ag by means of membrane-anchored Ag-receptor molecules: Ab molecules on B cells and T-cell Ag receptors (TCRs) on T cells. Each B cell and each T cell specializes in making Ag-receptors of a single specificity. Cells with identical Ag receptors belong to the same clone of cells. At the stage of mature B cells or mature T cells, each clone consists of only a few or a few dozen cells. However, since an organism has ~106 to 109 such clones with differing specificities, the organism has the capacity to recognize most foreign protein Ags.
Clonal Activation of Lymphocytes
When a mature B cell or mature T cell encounters and binds a threshhold dose of Ag molecules through its Ag-receptors and also receives other appropriate signals, it will proliferate and differentiate. This process ultimately yields an effective quantity of the Ab-secreting plasma cells or effector T cells, as well as a bank of temporarily dormant memory B cells or memory T cells that can be fully activated later. In this manner, an Ag can trigger a response limited to the given Ag and provide a cellular basis for memory. The Abs secreted by a clone of terminally-differentiated plasma cells all have the same specificity as the membrane-anchored Abs that served as Ag-receptors on the parental B-lymphocyte. For most complex protein Ags, many different B-cell clones within the animal's repertoire respond; i.e., the Ab response is polyclonal.
Ag Recognition Differs between B cells and T cells
B cells and T cells recognize Ag in different ways; this dual recognition system for TD Ags ultimately promotes the formation of B-cell-T-cell conjugates that are critical for the Ab response to TD Ags. B-cell Ag-receptors (or Abs) react with antigenic sites (epitopes) on the surface of the intact Ag molecule, whereas TCRs necessarily react with peptide fragments of a protein Ag. However, TCRs do not react with free peptide; instead they react with a complex of the Ag-derived peptide and a member of the cell-associated MHC (major histocompatibility complex) family of molecules. So-called "professional Ag-presenting cells" can internalize exogenous protein Ags, process them into peptides, associate the appropriate peptides with MHC molecules, and then display the peptide MHC complexes on the surface of the cells; they "present" the Ag-derived peptides to T cells. MHC molecules are broadly specific; each of the several varieties within a cell can bind any of several hundred or several thousand peptides (one at a time). However, only peptides with selected sequence motifs can bind with high affinity. Usually, but not always, a foreign protein will have one or more peptides that can bind one of the cell's MHC molecules with high affinity. It is important that a protein Ag (a TD Ag) have both B-cell epitopes and T
Cell-Cell Interactions and Cell Activation in the Humoral Immune Response
A snapshot of cell activation relevant to production of Abs directed to protein Ags is useful for understanding strategies for immunization and sites of action of adjuvants. The snapshot is necessarily an oversimplification, but serves as a practical guide (Figure 3). Two helpful paradigms are: (1) activation of cells is regulated, and each step in the pathway towards activation of B cells provides an opportunity for enhancing up-regulation of the response (by adjuvant components) or diminishing down-regulation; and (2) each lymphocyte needs two sets of signals to become activated. The first signal is delivered by Ag and usually involves engagement of a co-receptor in addition to the Ag-receptor on the lymphocyte. The second set includes costimulatory and differentiation signals delivered by: (1) direct contact with another cell through receptor-ligand pairs, and (2) cytokines (soluble polypeptide factors, variously called lymphokines, monokines, or interleukins). Activation of at least three cell types is required.
1. Dendritic cells and/or macrophages, called professional Ag-presenting cells (APCs), usually initiate the first events in the humoral immune response upon encountering a protein Ag. These cells internalize protein Ag molecules into an endosomal compartment, process (degrade) them to peptides, associate one or more of the peptides with MHC Class II molecules, and express the MHC Class II/peptide complexes on the cell surface. In this manner APCs present the processed protein Ag (peptide) to T cells. For the APCs to be effective in Ag presentation to T cells, they must present an effective number of peptide/MHC complexes on their surface (>100) and they themselves need to be in an activated state. Full activation of the APCs may require their sensing some bacterial product (endotoxin or peptidoglycan) or some host tissue product generated during tissue damage (an alarm signal) that leads to high-level expression of cytokines, cell-surface costimulatory molecules, and MHC Class II molecules by the APCs.
2. Clones of Ag-specific T-helper cells (CD4+ TH cells) are the next to be activated; they conjugate temporarily with the activated APC through TCR recognition of the peptide/MHC Class II complex displayed on the APC's plasma membrane (Signal 1 for the T cell). CD4 molecules on the TH cells, called co-receptors, help recognize MHC Class II molecules on the APC and help receive Signal 1. The conjugated TH cells then receive a costimulatory contact signal (Signal 2 for the TH cell) from the activated APC, leading to activation and proliferation of the CD4+ TH cells. Cytokines, such as IL-1, made by the activated APC may serve to deliver part of the costimulatory signal that amplifies the CD4+ TH cells. IL-2, a cytokine secreted by activated TH cells may behave in an autocrine manner, helping to further amplify the Ag-activated TH cells.
3. Finally, clones of Ag-specific B cells become activated. They receive Signal 1 when they encounter and bind the Ag by means of their Ag-receptor, i.e., membrane-anchored Ab. Signal 1 is much more effective if the co-receptor is also engaged; this can be accomplished by complement fragments deposited on the Ag. The B cell will internalize some of the Ag molecules, process them to peptides, associate one or more of the peptides with MHC Class II molecules, and then display the peptide/MHC Class II complexes on its surface. At this point the activated, Ag-specific TH cells will conjugate with the B cells through the surface peptide/MHC Class II complexes. The conjugated TH cells will deliver Signal 2 to the B cell, in part by receptor-ligand contact signals, and in part by cytokines that the activated TH cell secretes. Because B cells use the peptide/MHC Class II complexes to initiate contact with Ag-specific activated TH cells, and because the peptide fragments derive from the internalized Ag, it is usually necessary to have the B-cell epitopes and the T-cell epitopes on the same Ag molecule.
The activated B cells will proliferate and differentiate. Differentiation may proceed along two paths: (1) to Ab-secreting plasma cells, or (2) to memory B cells. Usually both plasma cells and memory B cells develop in response to T-dependent Ags, but sometimes memory can be developed without an apparent Ab response, and vice-versa. Cytokines delivered by the activated TH cells to the B cells will influence the classes or subclasses of Ab made during the secondary response. The kinds of cytokines delivered by an activated TH cell reflect its own earlier differentiation pathway, which may have been influenced by the type of APC, the local cytokine environment, the type of adjuvant, or all of these factors. TH1 and TH2, two subpopulations of CD4+ TH cells identified in the mouse, each secrete characteristic sets of cytokines. In the mouse, TH1 cells bias the secondary immune response towards cell-mediated immunity and the IgG2a subclass of Ab, while TH2 cells bias the response towards humoral immunity and IgE plus the IgG1 subclass of Ab. There is indirect evidence for the existence of such TH cell subpopulations in other mammalian species as well.
The Role of Lymphoid Tissues in the Humoral Immune Response
The cell-cell interactions generally take place in secondary lymphoid tissues (lymph nodes and spleen), first in the T-cell rich areas to yield activated TH cells, memory TH cells, and plasma cells responsible for the primary serum Ab response (predominantly IgM), and later, in the B-cell laden follicles of lymph nodes and spleen to yield, in part, memory B cells and, in part, developing plasma cells responsible for the secondary serum Ab response (predominantly IgG). Primary follicles that are seeded with a few activated B cells become germinal centers where the activated B cells undergo extensive proliferation and subsequent differentiation. In the germinal centers of mammals (and perhaps in chickens as well), V-region genes for Ab molecules undergo somatic hypermutation, a process that can generate some B-cell subclones that bear surface Ab molecules of higher affinity. This process of hypermutation is not fully understood, but it is known that it can result in Ab molecules that have some minor changes in structure in their Ag-combining sites. In any case, the highest affinity clones appear to be selected for survival. Somatic hypermutation and affinity selection work together over a period of time to increase the average affinity of the pool of Abs made by the organism. This process, referred to as affinity maturation, may increase the average affinity of the pool of Abs by more than 1000-fold. Thus the germinal centers are extremely important for the quality of the secondary immune response.
The affinity selection process and the maintenance of B-cell memory appear to be dependent on having some intact Ag retained in the germinal centers (reviewed by Burton et al., 1994; Thorbecke et al., 1994). The Ag-retaining reticulum of the germinal centers is comprised of specialized follicular dendritic cells (FDCs) that capture intact Ag in the form of Ag-Ab complexes by means of complement receptors or Fc receptors or both. Ag may be retained in the germinal centers for a matter of months to years, but will be diminished with time. After the germinal center reaction is well established, some of the memory B cells and memory T cells generated in the draining lymph node will circulate and enter other lymph nodes where they may be activated by a second dose of Ag. However, after a prolonged time without renewed Ag exposure, only the lymph node (or nodes) that generated germinal centers will contain the memory B cells. During the active secondary response, the majority of the developing plasma cells will leave the lymph nodes and take up residence in the bone marrow where their life expectancy is a few weeks to a month. Consequently, a prolonged Ab response requires continuous generation of new plasma cells by the lymph nodes, the spleen, or both.
GENERAL PROPERTIES OF ADJUVANTS
An adjuvant may be broadly defined as any substance that improves the immune response to an Ag. The term is generally restricted to substances that enhance the immune response when administered with the Ag, but it may be used in a broader sense to include immunostimulatory substances administered separately from the Ag. Adjuvants may function in various ways and, consequently, the variety of substances purported to have adjuvant properties is quite large (Warren and Chedid, 1988). Adjuvants used in human and veterinary vaccines for protective immunizations are more limited in scope and dose than those used for hyperimmunization of laboratory animals for Ab production. However, some of the latter adjuvants can also produce an exaggerated inflammatory response in the animal so that responsible use of adjuvants requires attention to their potential adverse side effects.
An adjuvant may function as an Ag-depot-forming substance, a delivery vehicle or inert carrier, an immunostimulator/immunomodifier (able to stimulate cells of the immune system or modify immune cell activation), or a combination of these. A single substance in an adjuvant may perform more than one of these functions, so the functional classification is often vague (Altman and Dixon, 1989; Kaeberle, 1986). Most, but not all commonly used adjuvant formulations for Ab production form an injection-site Ag depot, which provides a protected reservoir of Ag for slow release to draining lymph nodes. This process helps promote formation of memory cells and prolonged Ab responses (months for some adjuvants) without the need for repeated Ag injections. Also, injection-site depots of Ag and adjuvant can induce the formation of granulomas that may, in some cases, become additional sites of Ab production (Askonas and Humphrey, 1958; Leskowitz and Waksman, 1960). Emulsion adjuvants (such as Freund's adjuvants), alum, solid phase adsorbents (such as nitrocellulose), and some entrapping or encapsulating materials for Ag possess depot functions.
Many of the depot-forming materials may also be considered as vehicles or inert carriers of Ag as they often help direct Ag to secondary lymphoid tissues and they serve to concentrate or aggregate the Ag for its efficient delivery to relevant cells of the immune system. Particle size, or droplet size for oil-in-water emulsions or water-in-oil-in-water emulsions, influences Ag uptake by lymphatics and Ag-trapping by lymph nodes (surface retention by dendritic cells and phagocytosis by macrophages). Particles between 10 nm and 10 Tm readily enter the lymphatics but not the blood vascular system. Particles in the range of a hundred nm to a few Tm (such as bacteria) are readily phagocytosed by macrophages, and then processed for Ag presentation. Many vehicles or carriers may be classified as surfactants (Woodard, 1990); these substances have adhesive properties for protein Ags and for some immunostimulatory molecules as well, and may themselves have immunostimulatory properties. A surfactant's ability to concentrate protein Ag molecules on a surface along with the creation of a large surface area of Ag in the emulsion adjuvants, promotes efficient delivery of Ag and stimulators to the relevant cells of the immune system. Some vehicles or carriers may not have a significant depot effect (such as ISCOMs and zinc proline) while others are not inert carriers but combine the carrier function with immunostimulatory properties (such as Freund's adjuvants and alum).
Immunostimulators and immunomodifiers help activate the cells of the immune system, either directly or indirectly. They may enhance cell proliferation , cell differentiation, or both. They may bias differentiation of naive CD4+ TH precursor cells into TH1 or TH2 cells and thereby bias TH-cell cytokine , which ultimately affects the balance of cell-mediated immunity and humoral immunity as well as the class or subclass (isotype) of Ab produced. Also, they may promote activation of complement by the alternative pathway, which can have several immunostimulatory consequences. First, some of the generated complement fragments may attach to the Ag or inert carrier and target the Ag to macrophages through the macrophage's complement receptors, which ultimately enhances Ag presentation to T cells. Second, some of the generated complement fragments can deposit on the Ag or inert carrier and engage the B-cell co-receptor, which leads to significant enhancement of B-cell stimulation by the Ag (Fearon and Carter, 1995). Third, complement fragments deposited on the Ag will aid in Ag-trapping/Ag-retention by follicular dendritic cells (FDCs) in lymphoid follicles. Ag-retention by the FDCs appears to be necessary for memory in the B-cell compartment, for secondary Ab responses, and for affinity maturation of the Ab response. And fourth, the complement cascade generates other immunostimulators and immunomodifiers through activation of macrophages, platelets, basophils, and mast cells. Conversely, immunostimulators and immunomodifiers may also block the normal feedback suppressive activities, and thereby increase Ab production. Chemicals in the immunostimmulator/immunomodifier functional group are very diverse and may include mineral salts, surfactants, peptides, and cytokines or cytokine fragments.
A frequent site of action of immunostimulators is on the macrophages. These cells apparently have a number of cell surface receptors capable of responding to components of bacteria or yeast or damaged host tissue with an alarm signal, setting in motion their own activation and the subsequent activation of other cells needed for Ab production (Janeway, 1993). Some adjuvants contain killed bacteria for this purpose (such as Mycobacteria sp. in Freund's complete adjuvant, or Bordetella pertussis), but whole bacteria frequently produce intense inflammatory reactions. In recent years, identification of active components of bacterial and yeast cell walls has led to use the of purified active components and analogues thereof in an attempt to optimize immune enhancement while minimizing the generalized inflammatory responses. Immunostimulators related to components of cell walls include trehalose dimycolate (TDM) from the "cord factor" of mycobacteria; MDP (muramyl dipeptide) from the peptidoglycan fraction of cell walls, or less inflammatory analogues of MDP; monophosphoryl Lipid A (MPL), modified from lipopolysaccharides of gram-negative bacteria) (Rudbach et al., 1988; 1995); and LTA (lipoteichoic acids) from the cell walls of some gram-positive bacteria (Kotani, 1992). Not all immunostimulators work for all species (or even all inbred strains) equally well (Byars and Allison, 1995; Rudbach et al., 1995), perhaps because different species have evolved different alarm signal receptors on their phagocytes and other cells to identify and combat the most dangerous microbes in their own realm.
CHOICES AND DECISIONS IN MAKING POLYCLONAL ANTIBODIES
Major decisions in making polyclonal Abs center on choice of the species to be used for immunization and on design of an immunization protocol. The immunization protocol requires decisions on the form of the Ag-of-interest (such as size and state of nativity), the quantity of Ag, the route of injection, number and distribution of injection sites, the frequency of Ag injections, the particular adjuvant, the quantity of adjuvant, and the Ag:adjuvant ratio. Optimization of the protocol for Ab production is empirical for each Ag/adjuvant combination in each species, but some guidelines can be used to design protocols to produce high quantity/high quality polyclonal Abs without lengthy trials. Investigators must always consider the potential of adjuvants to cause pain and distress to the recipient and must maintain a concern for the welfare of the animal in designing their immunization protocols.
SELECTION OF ANIMALS FOR POLYCLONAL AB PRODUCTION
Species
In choosing a species for making polyclonal Abs, an investigator must consider: (1) the quantity of Ab or antiserum needed, (2) the phylogenetic relationship between the recipient and the donor of the protein Ag, and (3) the character of the Abs made by the recipient species (such as complement-fixing character). A number of vertebrate species (ranging from large farm animalsparticularly sheep, goats, and horses--to small laboratory rodents, chickens, frogs or fish) have been used over the years, each species offering some advantage for the particular study. The mammalian species used most frequently for polyclonal Ab production in the laboratory setting are listed below (Table 1) along with their approximate blood volumes and suggested single sample bleeding volumes.
For producing a polyclonal antiserum to a protein Ag, one generally chooses a donor/recipient pair whose members are not closely related on a phylogenetic basis. A greater phylogenetic distance is predictive of a greater number of amino acid sequence differences between homologous proteins, which translates to a more diverse and potentially higher-titer serum Ab response. This is because, in general, the animal's adaptive immune system responds to individual antigenic sites (epitopes) on a foreign substance and is tolerant to self; more epitopes yields more Abs. On the other hand, the need for a highly specific antiserum directed to a limited number of epitopes calls for immunization of a closely related species, or even genetically divergent members of the same species for production of anti-allotype antisera. Mammals hyperimmunized with a large complex protein from a phylogenetically distant species may make as much as 10 mg/mL of specific Ab, whereas animals hyperimmunized with a protein allotype (from a genetic variant within the species) may make less than 1 mg/mL of specific Ab.
Rabbit. The rabbit has been the single most used species for antibody production because is a convenient size, is easy to handle and bleed, has a relatively long life span (5-8 years), and produces adequate volumes of high-titer, high-affinity, precipitating antisera (Stills, 1994). The rabbit, a lagomorph, has diverged significantly from rodents, such that rabbit is usually suitable for making Abs to rodent proteins as well as to human proteins. The rabbit has a single IgG class (no subclasses) (Table 2). This means that the IgG Ab is generally predictable in its Fc region-determined functional character (such as complement fixation and Protein A binding), although quantitatively minor variations in function may result from differential glycosylation of IgG molecules (Fanger and Smith, 1972). The existence of major allotypic differences of rabbit immunoglobulins allows preparation of anti-allotype Abs that can be useful as antiglobulin reagents in immunoassays. Fc, Fab, and F(ab')2 fragments can be prepared from rabbit IgG with relative ease (Harlow and Lane, 1988).
Chicken. The chicken provides a distinct advantage for making polyclonal Abs to mammalian proteins because of its phylogenetic distance from mammals (reviewed by Larsson et al., 1993). The chicken may be particularly useful when (1) the investigator wishes to use an intracellular mammalian protein as Ag (Gassmann et al., 1990) (because homologous intracellular proteins tend to be more conserved in sequence between mammalian species than do homologous extracellular proteins), or (2) the cross-reactivities or Fc-dependent effector functions of mammalian immunoglobulins have the possibility of interfering with the intended Ab use (Larsson et al., 1993). While the properties of chicken IgY (also called IgG) may be advantageous in some immunoassays, in others it is a disadvantage. Chicken IgY does not activate mammalian complement component C1 (although it does activate the chicken homologue), nor does it react with bacterial Protein A or Protein G, mammalian Fc receptors, or mammalian rheumatoid factors (Larsson et al., 1993). Chicken IgY does not usually behave as a precipitating Ab in physiological saline, but it can be effectively used for precipitation in a neutral high salt solution (1.5 M NaCl) or with an acidic (pH 5), low ionic strength salt solution (Gallagher and Voss, 1970). Shimizu et al. (1992) reported that chicken IgY is less stable under extreme conditions than rabbit IgG Ab in that it denatures with loss of Ag-binding activity below pH 4.0, or above 62.5°C. However, IgY is stable within the egg for months, and purified IgY can apparently be stored stably for years in neutral buffers in the cold (Larsson et al., 1993). Fc and monovalent Fab or Fab' fragments have been prepared from chicken IgY, but bivalent fragments have not (Akita and Nakai, 1993a).
The fact that chickens transfer a significant quantity of IgY (IgG) to egg yolk, and the development of relatively efficient methods for purifying IgY from egg yolk (Akita and Nakai, 1993b; Landon et al., 1995; Svendsen et al., 1995) have made production of chicken Abs an increasingly acceptable, if not preferable, option for many studies. Harvesting the Abs from eggs obviates the need for an invasive collection procedure. Moreover, a week's worth of eggs can yield up to 10-fold greater quantity of Ab than the volume of rabbit blood collected from a weekly bleeding (Gassmann et al., 1990). Inbred strains of the species are available (Abplanalp, 1979), which may yield more reproducible immune responses than outbred stocks.
Mouse. Inbred strains of mice are available, which allows the investigator to select the highest responders from among a number of strains tested with the Ag-of-interest. Although mice are used frequently for production of monoclonal Abs, their small size has generally been considered an impediment to collection of enough serum for harvesting adequate quantities of polyclonal Abs. However, the use of mice for polyclonal Ab production may be particularly advantageous when the quantity of Ag is limited to a few (g or when only a limited amount of the Ab is needed. On the other hand, mouse size may no longer be a major drawback in terms of quantity of Ab that can be harvested, as polyclonal Abs can be harvested from ascitic fluid. A number of methods have been developed for producing ascites in mice (Cartledge et al., 1992; Kurpisz et al., 1988; Lacy and Voss, 1986; Mahana and Paraf, 1993; Maurer and Callahan, 1980; Tung, 1983) with variable success. Karu has recently developed a method using T-180 sarcoma cells to induce ascites in polyclonal Ab-producing Swiss Webster mice, such that 10-40 mL of Ab-containing ascitic fluid can be collected from a single mouse (Karu, 1993; Ou et al., 1993; Sartorelli et al., 1966). This method appears to be superior to other reported methods for ascites production, being relatively reliable, giving high yields, and not requiring repeated i.p. injections of Freund's adjuvant or pristane.
Rat, Hamster, Guinea Pig. Rats, hamsters, and guinea pigs are used less frequently than rabbits for polyclonal Ab production as these species usually offer no significant advantage over the rabbit. Blood volumes are smaller than that of the rabbit and collection of significant quantities of blood generally requires anesthesia and cardiac puncture (Flecknell , 1995). However, some situations arise in which these species have an advantage.
The rat is particularly useful for making IgG Abs of restricted specificity to mouse proteins and is the species of choice for IgE Ab production (Garvey et al., 1977). Additionally, the availability of inbred strains of rats (Nomura and Potkay, 1991) offers some advantages when the rat is compared to the rabbit. Ab production-limits imposed by blood-sampling limits may be overcome to some extent by harvesting Abs from ascitic fluid. Douglas et al. (1979) reported a method using an initial intraperitoneal injection of Freund's complete adjuvant with subsequent injections of Freund's incomplete adjuvant for ascites production in female rats of various stocks and strains. The success rate was somewhat variable in outbred stocks, poor in males, but reliably yielded ~60 mL/female Lewis rat in 10 days. A disadvantage of this species is that some rat strains, and males in particular, are prone to develop adjuvant arthritis upon injection of Freund's complete adjuvant (Cohen, 1992; Petty et al., 1989).
Hamsters and guinea pigs are phylogenetically more distant from the mouse than is the rat and thus can be used to raise anti-mouse protein Abs when such Abs cannot be readily produced in the rat or when Abs broader in specificity than those raised in the rat are needed. Many species of hamsters exist. When reference is made to "the hamster" in the United States, it is likely to refer to Mesocricetus auratus, otherwise known as the golden or Syrian hamster (adult weight = ~80-120 g) (Van Hoosier and McPherson, 1987). However, the Armenian hamster, Cricetulus migratorius (adult weight = ~40-80 g), is frequently the species used for production of hamster anti-mouse Abs, especially when the interest is in making monoclonal Abs (Sanchez-Madrid and Springer, 1986). This is because very stable hamster-mouse interspecific hybrids can be made between Armenian hamster B-lineage cells and mouse myeloma fusion partners.
Guinea pigs are used less frequently now than in the past for production of Abs. Guinea pigs were found to make superior responses to porcine insulin relative to rabbits, presumably because of a significant difference in amino acid sequence of their own insulin (Odell et al. 1972). Guinea pigs can make Abs that are excellent in complement-fixation and precipitin tests when they are immunized with Freund's complete adjuvant, but they make poor precipitins with poor complement-fixing ability when immunized with Freund's incomplete adjuvant (Garvey et al., 1977). This information can be interpreted based on the TH1/TH2 model of immune response regulation in the mouse as follows: Freund's complete adjuvant effects a predominant TH1 response in the guinea pig leading to a predominant expression of the IgG2 subclass, the complement-fixing subclass of IgG that also has better precipitating quality. Although ascites can be induced in guinea pigs (Tung, 1983), the described approach requires multiple intraperitoneal injections of Freund's complete adjuvant and is seldom used for Ab production in this species.
Farm Animals. The larger species--goats, sheep, and horses--are used primarily when large volumes of antisera are needed. These species have a long life span, are relatively easy to handle, and can be bled from a jugular vein without anesthesia. They are also expensive (particularly if multiple animals are immunized) and require special housing facilities, which can be prohibitive to individual investigators. However, horses, sheep, and goats are frequently used for commercial production of antisera. Historically, horse Abs have been used clinically as therapeutics for toxins and venoms, and are still so used today (Theakston and Smith, 1995). The "flocculating character" (precipitation occurs within a limited molar ratio of Ag to Ab) of horse IgT Ab (also called IgGt), an IgG-like subclass that predominates in the anti-toxin and anti-venom specificities, may minimize host reaction to foreign passive Ab. Also, horse antisera have been used for detection of precipitin bands in immunoelectrophoretic assays of serum proteins because the horse Abs tend to give good resolution of preciptin bands in a multicomponent system. Their resolving power is due to their tendency to resolubilize precipitated immune complexes in Ab excess in a manner to form thin, well-separated precipitin bands (Hurn and Chantler, 1980). A disadvantage of the horse is its intolerance to oil-based adjuvants; immunizations with less effective adjuvants often require large quantities (100s of mg) of Ag. For this reason among others, some antivenoms have been developed in sheep and other species in recent years.
Age, Gender, and Number
In general, young adult animals of the species are better Ab producers than older animals and they can be expected to produce Abs over a longer period of time. By the time of puberty, the immune system will have matured but will not have accumulated competing environmental Ags and memory cells to the extent of older animals. Also, levels of potentially-interfering, passive (maternal) IgG Ab obtained by the animal in utero, in ovo, or neonatally will have declined to insignificant levels in all but the small, early-maturing species. In the rabbit, for example, passively acquired maternal IgG usually declines to undetectable levels between 8 and 12 weeks, whereas the young rabbit's IgG synthesis rises sharply between 4 and 8 weeks and approaches 80% of adult levels between 8 and 20 weeks (Dray, 1971). When testing groups of 2-, 4-, 8-, 12-, or 16-week-old mice for Ab response to one Ag/adjuvant combination, Hu et al. (1990a ) found the best responses in the groups of mice first immunized at 8-weeks and 12-weeks of age. The following ages (~early puberty) are the earliest ages usually used for initiation of immunization for Ab production: mice, 6 weeks; rats, 6-8 weeks; rabbits, 3-5 months (Harlow and Lane, 1988); goats, 6-7 months; sheep, 7-9 months (Morrow, 1980); chickens, 16-20 weeks (Landon et al. 1995; Lewis and Perry, 1994). In most cases, the immunized animals will have matured another 4-8 weeks before Abs are harvested.
Immune function apparently peaks at puberty and declines gradually with age thereafter. Involution of the thymus and a decline of regulatory T-cell function appear to be major factors for the decline (Hirokawa et al., 1994; Miller, 1992). Cell-mediated immunity declines earlier than humoral immunity (Maletto et al., 1994). Although serum immunoglobulin levels may be elevated in older humans and animals, an ability to mount primary responses is diminished (Paganelli et al., 1994). Also, germinal center formation is defective in aging mice, due in part to poor Ag localization and retention by the FDCs (Thorbecke et al., 1994), and thus secondary immune responses are also compromised. The gradual decline of immune function does not preclude use of mature animals for Ab production, but use of older animals may lead to a lower titer or a response of lower diversity. In studies comparing the Ab responses to trypanosome Ag in mice of various ages, the responses in mice first immunized at 9 months and at 12 months of age were ~25% and ~40% lower, respectively, than those found in mice 3 or 6 months of age (Maletto et al., 1994). Hirokawa et al. (1994) reported that mice 9 months of age or older showed a significant decline in the Ab response to sheep red blood cells, with the response at 24 months of age being only ~25% of that found at 6 months of age.
Although members of either sex may be used quite satisfactorily, females are sometimes preferred for one or more reasons. Obviously hens are advantageous for production of chicken Abs because of their ability to transfer large quantities of IgY to egg yolks and because Ab yields from egg yolks are better than Ab yields from blood. Females of some species are generally more docile and easier to handle and maintain (such as mice and goats). For larger farm animals, castrated males may be suitable. Also, adult females have been reported to mount more vigorous immune responses, both cellular and humoral, than males.
Sexual dimorphism in the Ab response has been documented in a number of studies (reviewed by Grossman, 1984, 1985, 1989). Females are apparently sensitive to lower doses of Ag for primary immunization, have significantly higher and more prolonged primary responses to equivalent doses of Ag, and may have somewhat higher and more prolonged secondary humoral immune responses. Terres et al. (1968) reported that female Swiss albino mice mounted a stronger and longer lasting primary Ab response to bovine serum albumin than male mice but the assay used did not permit direct comparison of quantities of Abs. Batchelor and Chapman (1965) compared hemagglutination titers of male and female (BALB/c x C57BL)F1 hybrids immunized with sarcoma cells sharing an Ag with the tested erythrocytes. Both primary and secondary responses were slightly more vigorous and of longer duration in the females. Blazkovec et al. (1973) and Orsini and Blazkovec (1974) compared the number of Ab-forming cells, which are also called plaque-forming cells (PFC), in spleens of post-pubertal male and female hamsters immunized with sheep red blood cells. In both the primary and secondary responses, females generally had approximately 1.5-3 times as many IgM- and IgG-PFC/spleen as males, although the differences were not generally significant when the results were expressed as PFC/106 splenocytes. This apparent discrepancy could be attributed to the larger spleens in the females. However, no major differences were detected in hemagglutinin titers or hemolysin titers of serum samples from males and females. The gender difference in serum Ab titer has not been noticed to be strikingly different in a number of studies. Hu et al. (1990a ) found no obvious difference in IgG Ab production between small groups of age-matched male and female mice, although the oldest group tested was 16 weeks of age at primary immunization. The lower response that has been observed in males is attributed, at least in part, to depressive effects of androgens (Batchelor and Chapman], 1965; Blazkovec et al., 1973), but estrogens are also believed to play a role in enhancement in the female (Grossman, 1989). In cases where mg/mL of Ab in plasma is higher in females, some of the gender effect on Ab yield may be offset by the greater size of males; an exception is the hamster as adult female hamsters may be larger than age-matched males.
It is always wise to immunize more than one individual of an outbred population or to immunize several animals from each of several strains when inbred animals are used, particularly when there is no previous experience with the immunogen in the selected species or strain (Coligan et al., 1995; Garvey et al., 1977; Harlow and Lane, 1988; Hurn and Chantler, 1980). Use of multiple animals guards against total failure due to events that lead to non-response and often provides a more diverse set of responses in terms of quantity of Ab, specificity, and affinity than possible from a single animal. Failure, low response, or epitope-response-bias within an individual animal may result from an ongoing infection or from its Ag-exposure history which may have rendered it tolerant to all or some epitopes on the Ag-of-interest. Failure can also result from the recipient's genetic background, especially from lack of an appropriate MHC molecule for presentation of an Ag-derived peptide to the T-cell repertoire. This inherited "defect" would affect all members of an inbred strain and may apply to random animals of an outbred population. Although MHC molecules are "broadly specific" in terms of the range of different peptides they may bind, there are cases in which no peptide from a protein can be effectively bound by the MHC molecules genetically endowed to a given animal. It is often in this way that Ag-responsiveness/non-responsiveness is linked to the genetic type of MHC. The reason for non-response or poor response is not always known, but the variability in response is well documented. Garvey et al. (1977) found that as many as 2 of 10 rabbits were low-responders to bovine serum albumin administered intravenously; Sigel et al. (1983) reported failure of response in one of two rabbits immunized with human prolactin and failure of a desired response specificity in two of three rabbits immunized with human growth hormone. Zanelli et al. (1983) found only 5 of 122 antiserum samples tested to be of sufficient titer and appropriate specificity to be useful for parathyroid hormone assays.
When low availability of Ag or total cost is a factor in the use of multiple animals, the investigator may take a sequential (but very slow) approach, immunizing one or two animals initially and proceeding to additional animals if the results are unsatisfactory. If after three or four animals have been injected with two or three booster immunizations and no significant Ab response has been obtained, an alternate protocol should be used, the species species or strain immunized should be changed, or both (Coligan et al., 1995; Hurn and Chantler, 1980). Harlow and Lane (1988) recommend a minimum of two rabbits and three to six mice or other small rodents; Hurn and Chantler (1980) recommend "at least four or five" animals; Maurer and Callahan (1980) suggest five or six rabbits.
Each antiserum should be kept separate until characterized. Quantity, affinity, specificity, and subclass profile of Abs may differ between serum samples; differences will exist between animals and even between different serum samples from the same animal, as the response may change with time. For example, Abraham (1974) found that affinity and titer changed significantly in anti-steroid sera collected over a period of a year from sheep immunized with steroid-protein carrier conjugates. Only selected samples may have the desired character. Ultimately a pool of the high-titer, high-affinity antisera may provide better results than a single antiserum if broad specificity antisera are needed or if a large quantity of antisera with consistent characteristics is needed.
THE IMMUNIZATION PROTOCOL
Size, State of Aggregation, State of Nativity of the Ag
Size, state of aggregation, and state of nativity of the protein Ag will affect the quantity and quality of Ab produced. Small polypeptides (<10 kDa) and nonprotein Ags usually need to be conjugated to a large immunogenic carrier protein to become good immunogens. This is especially true when the peptide-of-interest is short, perhaps ~15-20 amino acids, derived from a larger protein or by a biochemical synthesis. However, even proteins 30-50 kDa may be improved in their immunogenicity by cross-linking the molecules to each other or to an immunogenic carrier protein. The bigger, the better is a good rule of thumb for immunogenicity of proteins. Large proteins increase the likelihood of good T-cell epitopes and large proteins or aggregates are more likely to engage the Ag-processing/Ag-presenting cells with a threshhold dose of Ag. One function of an adjuvant is to help concentrate and aggregate the protein Ag; aggregation is especially important for peptides and small proteins when they are not conjugated to a large immunogenic carrier protein. It is important to distinguish the difference between the inert carrier functions of an adjuvant (concentration/aggregation and delivery of Ag) and the immunogenic function of an immunogenic carrier protein (provision of T-cell epitopes), which cannot be replaced by an adjuvant. Both functions are important. One must avoid administering soluble deaggregated proteins, particularly by the intravenous route, as this form of the Ag may be tolerizing rather than immunizing (Maurer and Callahan, 1980; McCoy et al., 1986).
For conjugating peptides to immunogenic carrier proteins, a bifunctional cross-linker such as m-maleimidobenzoyl-N-hydroxysuccinimide ester can be used; many other bifunctional cross-linkers are equally appropriate (Coligan et al., 1995; Harlow and Lane, 1988). Glutaraldehyde can be used for conjugating two dissimilar proteins or peptides or for cross-linking similar peptides to each other. Nonpeptide haptens (small non-immunogenic substances), such as steroids, may also be coupled to carrier proteins and the conjugate can be used to elicit T-dependent Ab responses to the hapten (Erlanger, 1980). Choice of the carrier protein has been discussed by Erlanger (1980). KLH (keyhole limpet hemocyanin) and bovine serum albumin are frequently used. KLH is preferred by many investigators because it is large and phylogenetically distant from vertebrate proteins; however, it is less soluble than bovine serum albumin when cross-linked. Poly-L-lysine, either as a linear molecule or as a branching structure, has been used in place of a natural foreign protein carrier to provide a backbone for attachment of peptides. Tam (1989) and Posnett and Tam (1989) describe a Multiple Antigenic Peptide (MAP) method in which a branching lysine structure is used as a core for peptide synthesis in a manner to create a molecule with multiple identical peptides or a molecule with multiples of two different peptides. This procedure creates a density of epitopes on the same molecule, which in theory should be helpful. However, the use of poly-L-lysine as a "carrier" or core for the antigenic peptides so as to minimize or eliminate production of Abs to a large foreign carrier protein has sometimes resulted in failure of production of Abs to the peptides as well (Erlanger, 1980), possibly because of the lack of suitable T-cell epitopes.
The epitope density for hapten-immunogenic carrier protein conjugates can influence the immune response. While studying steroid-protein conjugates as immunogens for anti-steroid Ab production, Erlanger (1980) found that a low hapten density resulted in low Ab titers, a moderate density was optimal, while a very high density was counterproductive (Erlanger, 1980). Coupling 10-20 small peptides to a foreign protein carrier such as bovine serum albumin would be in the range to be helpful. One is likely to increase the Ab response to the peptide while reducing the Ab response to the protein carrier since more of the carrier's usual B-cell epitopes would be inaccessible and unable to participate in activating B cells. Any Abs formed against the carrier protein from immunization with a hapten-immunogenic carrier protein conjugate can be easily removed from the resulting antiserum with a solid phase immunosorbent made from the carrier protein alone (Harlow and Lane, 1988; Coligan et al, 1995).
There may be occasions when an investigator wants to develop Abs to small peptides without introduction of a large, immunogenic carrier protein into the animal. Several strategies are available. If the small peptide-of-interest contains both T-cell and B-cell epitopes, then cross-linking the peptide to itself may obviate the need for an immunogenic carrier protein. Alternatively, the MAP system described above could be used. Recently some investigators have been successful in obtaining Ab responses to small peptides that do not contain integral T-cell epitopes. This was accomplished by use of liposomes for simultaneous delivery of two different peptides to relevant cells of the immune system: (1) the peptide-of-interest for Abs and (2) other small peptides capable of serving as T-cell epitopes (Gregoriadis, 1995; Gregoriadis et al., 1993). A similar approach could be used with the MAP system in which two different peptides are synthesized onto a core branched lysine. Even chemical cross-linking of two different peptides may produce a successful immunogenic polypeptide (Francis and Clarke, 1989; Zegers et al., 1993). To some extent, simultaneous delivery of two different peptides may be accomplished with the oil-in-water or water-in-oil emulsion adjuvants--especially with Freund's complete adjuvant that will itself contain some mycobacterial protein with T-cell epitopes (Atassi, 1986), but appropriate delivery with the oily emulsion adjuvants is apparently less efficient than with liposomes. One difficulty with use of small peptides as T-cell epitopes is that members of the population of animals to be immunized may not be universally capable of recognizing the chosen T-cell epitope(s). In this respect, genetic variation in the MHC genes among the members of an outbred population may be responsible for some non-responders.
In addition to considering size and state of aggregation of the Ag, the investigator must consider the possible states of nativity of the protein Ag at two important points: first, when the protein is used as the immunogen, and second, when the protein is used as the reactant with the elicited Abs. Antibodies raised against native proteins react best with native proteins, whereas Abs raised against denatured proteins react best with denatured proteins. This is because conformation of the polypeptide chain of the protein Ag is of paramount importance in defining the epitopes recognized by B cells or Abs and because conformation as well as solvent exposure of various aspects of the polypeptide chain differ between the native and denatured states of a protein. If, for example, the Abs are to be used on membrane blots where the proteins have been subjected to denaturing conditions, then Abs should be raised against denatured protein, but if the Abs are to be used to detect native protein or to block an active site of the protein, then the Abs should be raised against native protein. Adjuvants can make a difference in the nativity of protein Ags and therefore make a difference in the specificities of Abs that are raised. A preformed oil-in-water emulsion adjuvant with adsorbed Ag will maintain a greater degree of native protein structure of the Ag than a water-in-oil emulsion that requires vigorous mixing for preparation (Allison and Byars, 1991; Byars and Allison, 1995), although at least some Abs reactive with native protein are usually obtained with either of the two adjuvant forms (Kenney at al., 1989).
When a small peptide is used for the immunogen, an additional concern is whether the Abs raised to the peptide will react with the native parent protein. In general, the larger the peptide, the greater the likelihood that it will assume a conformation similar to that in the whole native protein. If the selected peptide is small and was derived from a site within the protein that is highly dependent on neighbor-influenced conformation, then the elicited anti-peptide Ab is unlikely to react with the native protein (but exceptions are known). Peptides derived from the N-terminal and C-terminal ends of a protein are more likely to assume conformations similar to those in the native protein than are peptides derived from inner regions of the molecule where complex folding patterns are likely. Thus, terminal peptides have a higher probability of eliciting useful Abs for study of native proteins.
Preparation of Ag Solutions for Injection
Aqueous Ags for vaccines should be prepared in a manner to prevent contamination and/or to eliminate contaminants, particularly extraneous bacteria or bacterial products that may cause sepsis or an extensive inflammatory reaction. Most protein Ags can be filter-sterilized through a microporous filter (0.22 mm pore size) of a type that has minimal adsorption of protein and minimal disruption of protein conformation. Low-binding membranes, such as those made of cellulose acetate, are generally suitable for this purpose. Measures should be taken at all stages of Ag preparation to minimize contamination of the preparation by bacterial endotoxin which is inflammatory and pyrogenic and may not be removed by filtration.
Quantity of Ag
Choosing the optimal quantity of Ag for the immunization protocol a priori is difficult because inherent properties of individual Ags vary and because Ag dose relates not only to species injected but also to other parameters such as accompanying adjuvant, route, and frequency of injection (Hu and Kitagawa, 1990; Hu et al., 1989a,b; 1990a,b; Cooper et al., 1991). In general terms, Tg to mg quantities of a protein Ag used in conjunction with an adjuvant are needed to elicit a high-titer serum Ab response in a laboratory animal (Coligan et al., 1995; Harlow and Lane, 1988). And although smaller doses may be used for smaller animals, Ag dose is not increased or decreased in proportion to body weight (Hurn and Chantler, 1980). A more sensible way to think of Ag dose is in terms of the number of lymphoid follicles to which the Ag will be distributed. For rabbit, a usual dose of a soluble protein administered with Freund's adjuvant is in the range of 50-1000 Tg; for mouse, 10-200 Tg; for goat or sheep, 250-5000 Tg (Harlow and Lane, 1988; Hurn and Chantler, 1980). For each Ag there is a dose range called the "window of immunogenicity" that is usually relatively broad; however, too much or too little Ag may induce suppression, tolerance, or immune deviation towards cellular immunity rather than an active humoral immune response for the given Ag (Maurer and Callahan, 1980). Compared to injection of Ag alone, injection of Ag plus an adjuvant generally permits use of a much smaller quantity of the Ag, greatly enhances the serum Ab response (Kaeberle, 1986), and minimizes the risk of inducing tolerance. Within the window of immunogenicity, with or without adjuvant, a larger Ag dose generally results in a greater Ab response (Cooper et al., 1991; Katsura, 1972) up to a point at which suppressive activities become exaggerated. However, one must remember that, in general, the goal for Ab production is to obtain a large quantity of high-affinity Abs during the secondary response, so that a high serum titer after primary immunization is not necessarily cogent. Hu and co-workers (Hu et al., 1989a,b) found that ~100 Tg of a hapten-carrier conjugate administered intraperitoneally in Freund's complete adjuvant gave a significantly greater Ab response at 4 weeks than did 10 Tg of the conjugate; but they also found that the 10 Tg dose of conjugate for primary immunization established a better platform for a 10 Tg booster in Freund's incomplete adjuvant at 4 weeks. The 10 Tg/10 (g protocol resulted in 2.1 mg/mL serum anti-hapten Ab at 6 weeks after primary immunization, whereas the 100 Tg/10 Tg protocol resulted in 0.8 mg/mL anti-hapten Ab. The conclusion that low priming doses may lead to greater secondary responses was also reached by Siskind et al. (1968) in studies with rabbit. Duration of the response may also be affected by dose. Jones et al. (1990) reported that whereas 0.5 g or 50 Tg effective peptide dose of a peptide-protein carrier administered to mice intraperitoneally resulted in almost equivalent Ab titers to the peptide shortly after the last booster, the higher dose gave longer lasting titers.
In cases where quantity of Ag is very limited, the protocol should be designed to optimize the delivery of the Ag to lymphoid tissues without concern for broad distribution. To this end, various investigators have used direct injection of Ag into the spleen or lymph nodes (Boyd and Peart, 1968; Sigel et al., 1983; Spitz, 1986) or implantation of membrane-adsorbed Ag under the spleen capsule (Nilsson and Larsson, 1992). The effectiveness is variable but some investigators have achieved Ab responses with ng-Tg quantities of Ag. On the other hand, some investigators would argue that intradermal immunization may result in more effective delivery of Ag to lymphoid tissues, as discussed below.
Production of the highest quantity of serum Ab is not always, or even generally, the best goal because a moderate quantity of high-affinity polyclonal Ab is usually preferable to a high quantity of low-affinity polyclonal Ab (Kenney et al., 1989). At high Ag doses, a disproportionate number of low-affinity B-cell clones will be activated, whereas at lower Ag doses only the high-affinity B-cell clones are activated. Essentially there is an inverse relationship between the dose of Ag and the affinity of the Ab elicited (Goidl et al., 1968; Siskind et al., 1968); that is, high-affinity clones out-compete low-affinity clones for a limited quantity of Ag. Because titer of the antiserum (titer is the reciprocal of the final dilution at which detection of activity is possible in a given immunoassay) is a function of both the quantity of Ab and the affinity of Ab, measurement of titer alone does not fully characterize the antiserum. However, a high-titer antiserum raised to low Ag doses and collected after a booster immunization several months into the response generally contains a large proportion of high-affinity Abs. Of course one must also be aware that titer is directly dependent on the sensitivity of the immunoassay, so that comparison of titers is only valid with data from the same standardized immunoassay. On the other hand, direct comparison of quantities of Ab can be made between studies when quantity of serum Ab is expressed in Tg/mL or mg/mL.
The average affinity of the Abs found in serum can be expected to increase over a period of several months after an animal has been immunized with a T-dependent Ag associated with a depot-forming adjuvant (Eisen and Siskind, 1964; Goidl et al., 1968; Siskind et al., 1968). Mechanisms within the organism's lymphoid tissues select and perpetuate the high-affinity clones in an environment of decreasing quantity of Ag over time. In mammals, this phenomenon of affinity maturation by selection is enhanced by the process of somatic hypermutation of the Ab V-region genes. During the stage of rapid proliferation of the activated B cells in germinal centers, mutations occur with a high frequency (~1 in 103 nucleotides per cell division) in the Ab V-region genes for both the light- and heavy-chains, but not in other genes. Subclones of B cells that have mutations resulting in high-affinity surface Ab are selected for survival, presumably by making contact with the Ag-retaining FDCs; low-affinity subclones do not survive. In poultry, the contribution of somatic hypermutation of Ab genes to affinity maturation has not been fully defined (Parvari et al., 1990), but in any case it has been shown that the average affinity of the Abs in chickens increases with time after immunization (Yamaga and Benedict, 1975), albeit, not to the extent found in rabbits.
One strategy for selection of relative doses of Ag for primary and secondary immunizations suggests a larger dose for the primary immunization. The rationale is that a large primary dose initially stimulates a broader base of B-cell clones for subsequent somatic mutation, and that selection processes in the face of decreasing available Ag will assure survival of only high-affinity clones. This strategy also assumes that memory cells can be stimulated with less Ag than needed for primary B- or T-cell stimulation. Another strategy suggests a higher Ag dose for the secondary immunization to compensate for immune-enhanced clearance of the Ag. However, it is probable that the Ag-Ab complexes formed with pre-existing Ab actually help deliver the Ag to the FDCs in the secondary follicles of the lymphoid tissue and thereby serve to enhance the Ag activity for secondary responses. One strategy for immunizing animals employs Ag-Ab complexes (preferably made in Ag excess) rather than Ag alone (Goudie et al., 1966; Harlow and Lane, 1988). In any case, small variations in the dose of the booster, whether halved or doubled relative to the primary immunizing dose, are probably not critical; however, there may be an initial effect on affinity. Most investigators use a booster dose equal to the priming dose or equivalent to one-half the priming dose (Hurn and Chantler, 1980). With respect to a given species, the "optimal" primary dose of a given Ag, if delivered with an effective adjuvant and by an effective route, is likely to be at the low end of the range suggested above. Good priming, even in the absence of a strong initial response, is like money in a savings account: A greater bank of memory cells will produce a bigger secondary response.
Route of Injection and Number of Injection Sites
The choice of Ag injection route is shaped to some extent by choice of species and choice of adjuvant, as well as by character, quantity, and volume of the Ag. Anecdotal data abound in the realm of the best route, but well-controlled studies are few--perhaps because the number of variables is so large. Possible and usual routes (Leskowitz and Waksman, 1960) are intravenous (i.v.); intramuscular (i.m.); subcutaneous (s.c.); intraperitoneal (i.p.); and intradermal (i.d.), which is also called intracutaneous. These routes are listed in probable order of increasing effect for soluble Ags administered without adjuvant (Hurn and Chantler, 1980). The order listed may change with particulate Ags and with the use of adjuvants. In addition to the above routes, intranodal or intrasplenic injection routes have been successfully used in cases of limited quantity of Ag. Footpad and intra-articular routes have also been successfully used, but suitability of these as well as intranodal and intrasplenic routes is usually assessed today on an individual basis in terms of the welfare of the animal when adjuvants are used. The effectiveness of the various routes apparently relates to the efficiency of delivery of Ag to lymphoid tissues (Hurn and Chantler, 1980).
Ideally, for an optimal Ab response, an effective form and dose of Ag plus adjuvant should be broadly distributed to lymphoid tissue. In theory, deposition of Ag in a greater number of lymph nodes should engage more of the organism's lymphoid follicles in generating cells to produce the Ab-of-interest, and in practice, this appears to be the case (Bennett et al., 1992; Vaitukaitis, 1981). The i.v. route has the potential for broad distribution of Ags; spleen serves as the primary filter and peripheral lymph nodes serve as secondary filters, in terms of lymphoid tissues. However, the i.v. route also has some limitations, as discussed below. Use of multiple primary immunization sites, as opposed to single site injection for i.d., s.c., or i.m. routes, will not only distribute the Ag to more lymphoid tissue and establish more germinal centers, but the lower dose of Ag per site will also help limit pathology from inflammatory reactions at the sites of Ag deposition (Stills and Bailey, 1991). However, no definitive study exists on the optimal number of sites for Ag administration by various routes. An upper limit will exist due to a finite number of pre-existing Ag-specific B and T cells. Jackson and Fox (p. 146 of this issue) and Jennings (p. 119 of this issue) review recommendations for injection routes and volumes for Freund's adjuvants. It is helpful for investigators to adopt a standard pattern of injection in order to carefully monitor each injection site for inflammatory reactions.
In apparent opposition to the above argument for use of multiple sites, Herbert (1978a) has indicated that "very slight, if any, advantage is gained by making multiple inoculations at many sites on an individual animal..." The discrepancy may relate, in part, to Ag dose and to adjuvant type and dose. Large Ag doses at multiple sites may surpass the capacity of the animal to respond, induce suppression, or both. A large dose of antigen or a large volume of an inflammatory adjuvant at a single site may actually help accomplish broad distribution as a result of inflammation, but at the expense of the welfare of the animal. Moreover, granulomas may form at injection sites in response to inflammatory depot-forming adjuvants and become a significant focus of Ab-forming cells. This involvement of granulomas in Ab production may raise the quantity of Ab produced but it is likely to lower the quality of the Ab due to lack of full participation of organized lymphoid tissue in the process of affinity maturation.
The distribution of Ag to the lymphoid tissue depends on the route of injection as well as the character and quantity of Ag and adjuvant. Although Ag administered by the i.v. route is broadly distributed, a significant proportion of the Ag may be entrapped by the liver, bone marrow, and lungs. The biodistribution of i.v.-administered Ag, if particulate, is primarily dependent on particle size and site of entry (Strand, 1986). Particles leave the vascular system primarily through discontinuous capillary beds found in liver, spleen, and bone marrow (O'Driscoll, 1992). Resident macrophages in these tissues will "clear" the Ag. Mechanical entrapment may occur in lung capillaries for particles larger than 7 Tm (Strand, 1986). Thus a portion of the Ag will not be available to cells involved in production of Abs, or may even have a suppressive effect. Moreover, the number of adjuvants that can be safely injected i.v. is limited, and unprotected Ag may be degraded more rapidly than adjuvant-associated Ag.
Other parenteral routes result in more limited distribution of macromolecular or particulate Ags as they generally bias delivery to draining lymph nodes through the lymphatics. This is because most vascular capillaries comprise a more closed system (continuous basement membrane and adherent endothelial cell junctions) than lymphatic capillaries (also called initial lymphatics), which have interrupted basement membranes and potentially large clefts (~10 Tm is not uncommon) between endothelial cells (Allen, 1967; O'Driscoll, 1992; O'Hagan et al., 1992). Ag uptake into lymphatic capillaries occurs primarily through intercellular clefts, although vesicular transport across endothelial cells or transport through transendothelial channels may also make some contribution. Supersaxo et al. (1990) showed that the proportion of substances absorbed by lymphatics rather than vascular capillaries increased with molecular weight (four substances between ~250 and 19,000 kDa were tested). Molecules (or particles) less than 5 kDa (or particles <1 nm in diameter) are readily absorbed by vascular capillaries; molecules 20 kDa and above (or particles >~5 nm in diameter) are poorly absorbed, if absorbed at all into the vascular system (Ballard, 1968; Supersaxo et al., 1990). On the other hand, large molecules and particles, even red cells, may be absorbed by terminal lymphatics, particulary by the diaphragmatic lymphatics of the peritoneal cavity. These lymphatics maintain a very open structure due, in part, to the continuous muscular activity of the diaphragm.
Each of the transfer routes for tissue-deposited or cavity-deposited Ags may be influenced by local anatomy and local conditions (e.g., inflammation or muscle activity). Moreover, the size, charge, and lipid content of the Ag and adjuvant vehicle not only influence Ag uptake from interstitial sites, but these parameters will also influence retention of the Ag by the lymph nodes (O'Driscoll, 1992; >]O'Hagan et al.</A, 1992; Takakura et al., 1992). Larger particles (~400 nm) have better lymph node retention once lymphatic uptake has occurred, but smaller particles (~50 nm) have better uptake into the lymphatics. Cationic substances have better retention than anionic substances, and anionic substances have better retention than neutral substances. O'Hagan et al. (1992) have reviewed studies of lymphatic delivery of drugs by particulates, liposomes, and emulsions. Although most studies are focused on delivery of drugs for cancer chemotherapy, many of the same principles apply to Ags.
The i.v. route. As stated above, this route will deliver Ags primarily to the spleen in terms of lymphoid tissues and secondarily to lymph nodes, especially the bronchial lymph nodes as demonstrated by Askonas and Humphrey (1958) in rabbits injected with killed pneumococci and alum precipitated ovalbumin. Herbert (1978a) considered i.v. to be the route of choice for small particulate Ags such as virions, bacteria, or cells. In a study of Ab responses of rabbits over a period of approximately 3 months, Webster (1965) found that the i.v. route was superior to an i.p. or s.c. route for administration of an aqueous suspension of live influenza virus. In this same study, an initial i.m. immunization of rabbits with virus emulsified in Freund's adjuvant followed by two boosters of aqueous viral suspension administered i.v., i.p., or s.c. did not change the order of effectiveness found for the various routes with the all aqueous vaccines (i.v.>i.p.>s.c.). The i.v. route for the single emulsion water-in-oil adjuvants and large particulate or viscous gel adjuvants is not advised because of the risk of embolism (Herbert, 1978a). For soluble aqueous Ags without adjuvant, the i.v. route does not establish an effective Ag-depot site outside of the lymphoid tissue, does not sustain a high-titer response when used for primary immunization, and may increase the likelihood of tolerance. Although primary immunizations by the i.v. route are not generally used for soluble protein Ags, secondary immunizations of aqueous Ag by the i.v. route may be quite effective after primary immunization with Ag associated with an effective adjuvant (Leskowitz and Waksman, 1960; Herbert, 1968); Ag-Ab complexes made in vivo will serve to aggregate the Ag. Secondary i.v. immunizations do, however, have a much greater potential to cause systemic anaphylactic reactions than do s.c. or i.m. injections. Adjuvants or vehicles that can be delivered i.v. include liposomes (Phillips and Emili, 1992; Gregoriadis, 1995)), ISCOMS (Coligan et al., 1995), dispersed alum (Herbert, 1978a; Katsura, 1972), and appropriately prepared water-in-oil-in-water double emulsions (Herbert, 1978a), although the latter two may cause dispersed inflammatory reactions in some non-lymphoid tissues such as liver and lung. Species in which i.v. injections are used are generally limited to those with easily accessible veins. Techniques for i.v. injections of various species have been described in a number of methods books (Flecknell, 1995; Garvey et al., 1977; Harlow and Lane, 1988; Herbert, 1978b).
The i.m. route. Herbert (1978a) described this as the route of choice for Freund's complete adjuvant (FCA) in that it provides rapid access to the circulation and lymphatics. The i.m. route for FCA is a subject of considerable debate and is discussed further by Jennings on p. 123 of this issue. The rapidity of absorption of Ag and the distribution between vascular circulation and lymph is likely to depend on the Ag and the adjuvant as well as the muscle site of injection. Skeletal muscle is well vascularized and rapid absorption of small molecules (<2 kDa) readily occurs through vascular capillaries, particularly with increased blood flow during muscle activity. Muscle is considered a good site for injection of small molecular weight drugs with irritant properties because the drug absorption into blood is rapid and the focus for inflammation is dispersed (O'Driscoll, 1992). However, large molecules and particulates are unlikely to penetrate vascular endothelium via intercellular junctions unless the capillaries are damaged or there is severe inflammation. For some large molecules or particulates, vesicular or channel transport of Ag across the vascular endothelial cells may occur, but this is not likely to be a major route. It is possible, though not well documented, that emulsion adjuvants with detergents may cause lesions in the vascular bed so that a portion of oil-adjuvanted Ag is directly distributed to the circulation. However, large molecules and particulates are likely to be absorbed primarily by the lymphatics with delivery to the draining lymph nodes. The lymphatics of skeletal muscle are confined to the fascial planes surrounding muscle bundles; no lymphatics are known to exist in the delicate sheath (endomysium) surrounding individual muscle fibers, such that lymphatic drainage of Ags deposited deep within large inactive muscle bundles is unlikely to be rapid (O'Driscoll, 1992). However, muscle activity--either active or passive--greatly enhances lymph flow from muscle.
An advantage of the i.m. route is that relatively large volumes of material can be accomodated in animals with large muscles. Many veterinary vaccines are administered by the i.m. route. However, when a large volume of Ag is injected into a muscle bundle, a portion of the injected Ag may come to reside in the interfascial planes between muscle bundles (Droual et al., 1990). Ag can be absorbed by lymphatics in this region but Ag and adjuvant can also spread along interfascial planes and establish contact with nerve bundles where pathology consequent to inflammatory processes may occur.
The s.c. route. This is a route of convenience for investigators, because the injection is easily administered in most species and relatively large volumes may be accommodated. However, it is not wise to administer large quantities of Ag or large volumes of inflammatory adjuvants to single sites (see Jennings, p. 123 and Jackson and Fox, p. 147, this issue). In general, this route provides for slow absorption of Ag, which occurs primarily through the lymphatics. The s.c. route is often used for booster injections of aqueous Ag when there is risk of anaphylactic reaction should Ag reach the vascular system rapidly. The rate of absorption of s.c.-deposited material is a function of the area contacted by the Ag, vascular blood flow (skin temperature), and of the activity of underlying muscles (as well as a function of the size and character of the Ag). Spread of s.c.-deposited Ag will vary with species injected and with site of injection. Ag will be dispersed over a greater area in animals with loose skin (such as the rabbit) than in animals with tight skin (such as the guinea pig); likewise, different sites within an animal will vary in skin tightness and Ag dispersal. Hyaluronidase included in the injected material will cause increased Ag-spreading (Goldstein et al., 1974). Inflammatory reactions resulting from emulsion adjuvants may also cause spreading.
A variation of the s.c. route involves implantation of a plastic perforated "golf" ball (similar to a whiffle ball) in a subcutaneous space and injection of Ag into this chamber. Granuloma fluid can be repeatedly withdrawn from the chamber to obtain the Abs. This procedure has been reported for rabbits, rats, sheep, and chickens with variable success (Clemons et al., 1992; Ermeling et al., 1992; Hajer et al., 1977; Hillam et al., 1974; Reid et al., 1992; Wolff et al., 1976). This method has the drawback of an initial surgical procedure and a 4-week to 6-week delay for healing and granulation tissue formation. However, it has the advantage of ease of repeated samplings with little or no insult to the animal. Sampling may remove some of the Ag, persistence of which seems to be necessary for continued Ab production. An adjuvant appears to be necessary unless the Ag has its own immunostimulatory properties. In one study, use of acrylamide-entrapped Ag was successful for immunization and boosting; immobilized Ags may offer the advantage of retention of Ag within the cavity. No study to date has reported on the quality of the Ab obtained by this method. It is unlikely that affinity maturation proceeds in the granuloma to the extent that it does in germinal centers of lymph nodes and spleen, so that high quality Ab may better be made by using a method that more fully engages the animal's secondary lymphoid tissues.
The i.p. route. For delivery of Ags via lymphatics, the i.p. route is efficient for several reasons: (1) access of Ag to lymphatics is not impeded by interstitium as the Ag is in a serous cavity, (2) the terminal lymphatics of the diaphragm--comprising the major route of lymphatic drainage of the peritoneal cavity--maintain large intercellular clefts, and (3) respiratory activity helps pump the lymph towards the draining mediastinal nodes (Allen, 1967; O'Driscoll, 1992) and subsequently to the thoracic duct and the vascular system. Parker et al. (1981) in studies with rats showed rapid uptake of liposomes into diaphragmatic lymphatics as evidenced by high levels of labeled liposomes in thoracic duct lymph. They also showed that liposome-associated drug were evidenced in three major lymphoid tissues, the thoracic nodes, renal nodes, and spleen, in decreasing order. The renal node uptake was attributed to retroperitoneal lymphatic drainage. The presence of liposomes in the thoracic duct indicates that the i.p. route can provide broad delivery of Ag, to spleen and other lymphoid tissues, if a portion of the Ag escapes retention by the thoracic nodes. Red blood cells introduced i.p. may pass directly into lymphatics, traverse the mediastinal nodes, and flow through the thoracic duct to the vascular system (Allen, 1967). Other particulate Ags probably exit almost exclusively through the lymphatics. However, when Freund's complete adjuvant is used, it is possible that some direct uptake of macromolecules by vascular capillaries may also occur due to microscopic tissue damage and excessive inflammation. The contribution of dendritic cells to the transfer of soluble Ag from the peritoneal cavity to lymphatics has apparently not been studied.
Major advantages of the i.p. route are: (1) relatively large volumes of inoculum can be accomodated; (2) adjuvants of many types can be accomodated; and (3) the Ag, when administered in sufficient quantity, is broadly distributed. However, because of rapid absorption of Ag from the peritoneal cavity and potentially broad distribution via the vascular system, booster injections of aqueous Ags by this route may cause anaphylactic shock. The i.p. route is the most frequently used route for immunization of mice; it is also frequently used for other small rodents, but only occasionally used for rabbits and larger animals. In mice, the i.d. and i.m. routes are possible, but technically difficult and they impose volume limitations.
The i.d. route. This route is favored by many investigators. Although vascular absorption from this site is generally low, lymphatic delivery is relatively rapid. Abundant Langerhan's dendritic cells in the skin are able to carry intact Ag as well as processed Ag to draining lymph nodes in a short period of time. Freund (1951) first reported that the quantity of Ag that exited from an i.d. site within 30 minutes of injection was sufficient to cause Ab production. The i.d. route was popularized after Vaitukaitis et al. (1971) reported successful production of Abs useful in sensitive radioimmunoassays by one-time, multiple-site immunization of rabbits with as little as 20 Tg of hormone-protein carrier conjugates. In this procedure, further described by Nieschlag et al. (1975) and Vaitukaitis (1981), fur was shaved from the back and from proximal areas of limbs of 3-month old rabbits and 2 mL of the Ag/FCA emulsion was injected i.d. into 30-50 sites. The investigators used high quantities of Mycobacterium tuberculosis in the FCA (and stressed the importance of the high concentration) and also injected Bordetella pertussis (which behaves as an additional immunostimulator with a bias towards a TH2 response) at a separate site. Peak titers were generally reached between 2 and 4 months after the primary immunization. Modifications of this procedure, some with booster injections and some without, were then used in other species, including rat (Hillier et al., 1975) and sheep (Scaramuzzi et al., 1975) to develop high-titer, high-affinity antisera suitable for use in radioimmunoassays. The rationale behind the method, as described by Neischlag et al. (1975), "is to reach as much lymphatic tissue with as little Ag as possible, assuming that low Ag concentrations reach only high affinity Ab-producing cells (and that) high concentrations may not only merely reach low affinity cells, but may also even suppress the high affinity cells."
The footpad (toepad) route. I.d. in theory, in practice this route is probably a mixture of i.d and s.c., and perhaps even i.v. in unskilled hands. Suggestive evidence for the i.v. component comes from a study showing systemic anaphylaxis in rats minutes after subplantar booster immunization (Levine and Saltzman, 1994) and anecdotal reports of anaphylaxis after booster immunizations by the footpad route. The footpad is one of the most distal anatomic sites for delivery of Ag to the lymphatics, which thereby increases the probability of lymph node retention of Ag. The footpads are, however, weight-bearing surfaces of the animal and therefore particularly sensitive, especially for heavier animals and when inflammation of the tissue develops. This is a major concern when a depot-forming adjuvant is used. Rats and mice use their forepaws for food handling and thus these footpads should not be used when inflammation is likely. Use of the footpad route is discouraged for inflammatory adjuvants and for heavier animals, but, when used, inoculation is usually confined to a single foot (Herbert, 1978b). Rabbits should not be immunized in their "footpads" which are, in fact, not true footpads (Amyx, 1987).
The intranodal and intrasplenic routes. Both of these routes have been successfully used for direct delivery of small doses of Ag to the lymphoid tissues. Boyd and Peart (1968) and Sigel et al. (1983) elicited anti-peptide hormone Abs in rabbits injected in popliteal lymph nodes with Ag emulsified in FCA. However, discretionary use of adjuvants today suggests use of a less inflammatory adjuvant (Spitz (1986) used alum) in such a sensitive tissue, or simply aggregation or solid-phase immobilization of the Ag prior to its injection. Horne and White (1968), working with guinea pigs, found no advantage of intranodal immunization over footpad immunization with Tg quantities of Ag. Hurn (1974) has argued that direct injection of Freund's adjuvant into the lymphoid tissue often destroys the architecture of the tissue, and suggests that this may explain the variable success with this method. For mice, intrasplenic implantations of ng quantities of Ag coupled to Sepharose or adsorbed to nitrocellulose have been successfully used, but in some cases several booster implantations seemed to be necessary (Grohmann et al., 1991; Nilsson et al., 1987; Nilsson and Larsson, 1992). This procedure in mice, coupled with the subsequent production of ascitic fluid, may yield significant quantities of polyclonal Ab with approximately 1 Tg of Ag (Karu, 1993; Ou et al., 1993). Some investigators suggest that s.c. deposition of Sepharose-coupled or nitrocellulose-adsorbed Ag may be almost as effective as intrasplenic immunization in mice and rabbits (Coghlan and Hanausek, 1990; Nilsson et al., 1987).
Comparative studies on routes of immunization. The various sets of data comparing routes of immunization do not all support the same conclusion about the "best route". However, in general, the probable order of efficacy listed by Hurn and Chantler (1980) for aqueous soluble Ags--i.v.< i.m.< s.c.< i.p.< i.d.--may approximate reality even when adjuvants are used, except in the case of the i.v. route. The superior rank for the i.d. route can be understood on basis of efficient lymphatic delivery by Langerhan's dendritic cells as well as by broad distribution of Ag associated with multiple site injection if carried out by the method of Vaitukaitis (1981) and Vaitukaitis et al. (1971). The order of i.m., s.c., and i.p. may be interchangeable with various studies. The i.v. route is listed as the least effective because response to Ag without the use of an adjuvant (the more usual situation for the i.v. route) was compared to other routes with the use of adjuvant. On the other hand, Leskowitz and Waksman (1960) found the i.v. route to be superior to all others tested in rabbit, except for the toepad route, when bovine serum albumin, at relatively high Ag dose (2.5 or 25 mg bovine serum albumin) in FCA, was first administered i.v. in a dispersed water-in-oil-in-water emulsion (not a recommended procedure today). The superiority in the serum Ab response was seen after an i.v. booster of 10 mg bovine serum albumin in saline (given after ~5 months). This result was quite different from that found in the primary response where the i.v. route appeared to be the poorest of all routes tested. Comparison of Ab quantity 1 week after an aqueous i.v. booster gave the following order: s.c.< i.p.»i.m. < i.d. < i.v. << toepad, whereas prebooster showed i.v.<< i.d. << i.p.»toepad << i.m.»s.c. when measured at 6-9 weeks post immunization. Kinetics of the response were different for different routes; serum Ab quantity peaked at about 3 weeks for the i.v. route, at about 4-6 weeks for the i.d. and toepad routes, 6-9 weeks for the i.p. and s.c. routes, and 14-18 weeks for the i.m. route. (It should be noted that in this study, all routes, except i.v. and i.p., used four immunization sites, at the "four corners of the animal"). It is probably not meaningful to rely on Ab titers obtained after the primary response to judge efficacy, especially when using FCA, which is believed to bias the immune response towards a cell-mediated response (TH1 response). A booster immunization with Ag in Freund's incomplete adjuvant or with Ag in saline will move the response toward a dominant humoral immune response with balanced IgG isotypes.
Cooper et al., (1991) found the i.p. route in mice to be superior (~2X-8X) to the s.c. route, based on Tg/mL of serum IgG anti-KLH (keyhole limpet hemocyanin) Ab produced 1 week after booster (IgG Ab: ~100-200 Tg/mL). The experiments were done with suboptimal doses of KLH 1 Tg administered with Algammulin® as adjuvant for primary immunization and with no adjuvant for the booster at 2 weeks. Hu et al. (1990b) also found the i.p. route superior (~3.5X) to an s.c. route in mice in terms of IgG Ab (at 4, 6, and 8 weeks) when primary and secondary immunizations (booster at 4 weeks) were done with near optimal doses of a hapten-carrier (~10 Tg) emulsified in Freund's adjuvants (complete and incomplete). The IgG anti-hapten responses to the same Ag in saline (no adjuvant) administered i.v. or by intrasplenic injection were poor and negligible, respectively. On the other hand, Kenney et al. (1989) found the s.c. route superior (~1.5X-3.5X) to the i.p. route in mice in terms of both Tg/mL of serum Ab and affinity of Ab to human serum albumin (HSA). In these studies the HSA (~5 Tg) was administered with either SAF-1 or Al(OH)3/MDP as adjuvant and the Ab was measured after two booster injections. The superior response to the Ag administered by the s.c. route in the latter study could reflect differences in the kinetics of the responses since Ab response was assessed at a single time point, but it may also be due to other variables as well. The i.p. and s.c. routes are the two routes used most often in mouse; both the i.m. and i.d. routes are technically difficult in this species and accomodate only small volumes of Ag. Scaramuzzi et al. (1975) found the s.c. and i.d. routes of similar efficacy in sheep, when immunizing with a steroid-carrier protein conjugate (5 mg for the primary and 2 mg for booster injections at ~4 months) emulsified in FCA at 20 sites for the i.d route and 8 sites for the s.c. route. Serum Ab titers were tested intermittantly over a period of 167 days post primary immunization. Lader et al. (1974) compared i.d. (40 injection sites) and i.m. (4 injection sites) routes for primary immunizations for four different Ags in rabbits (6 in each group). FCA was used and the booster injections were by the i.m. route. No significant difference was found, although the affinity of the Abs from the i.m. group tended to be slightly higher.
Preferred routes. Results of the above studies indicate that no single route can be regarded as universally superior for the primary immunization. The variables involved are not completely understood. Choice of route may be based on choices of species and adjuvant. The choice of adjuvant may eliminate some routes. For example, the i.v. route with Freund's adjuvants causes severe inflammatory reactions in lungs and liver and is generally considered inappropriate. Some routes are technically or logistically difficult for some species, such as i.d. or i.m. for the mouse. Certain routes have become "preferred" for various species, although preference is often based on considerations other than largest Ab response.
For mice, Ag in an emulsion adjuvant is usually injected either s.c. (at the base of the tail or across the back), or i.p. The i.p. route is preferred when volumes over 100 TL are involved. Booster injections of Ag without adjuvants may be administered i.v. through the tail veins. For rats, any of the usual routes are used. For aqueous Ag solutions, the i.v. route, by tail vein, is feasible in young rats, but less feasible in older rats with thick and scaly tail skin. Guinea pigs may be injected i.m., s.c., i.p., or i.d., but they are seldom injected i.v., due to lack of a readily accessible vein. The penile vein can be used in male rats and guinea pigs, but anesthesia is required as the discomfort index is high. In general, use of this route is discouraged. Larger guinea pigs may have a suitable marginal ear vein. Rabbits may be easily injected by any of the routes, although i.p. is seldom used. For large farm animals, the preferred route is i.m. followed by s.c.; the i.d. route has been used in sheep in at least one study, but appeared to offer no distinct advantage over the i.m. route for the Ab response. Chickens are usually injected by an i.m. route (the pectoral muscles, lower leg muscles, or caudal neck), or s.c. route (in the dorsal caudal neck region, the axillary region, or the ventral junction of the thigh and abdomen). Intravenous and i.p. routes are used less often and "footpad" injection has been used occasionally.
The route for booster immunizations. Booster immunizations do not need to be administered by the same route used for the primary immunization. Boosting with aqueous Ag by the i.v. route is quite effective (Herbert, 1978a). This route involves more of the lymphoid tissue in producing Ab, even when the primary response was localized to regional lymph nodes following Ag administration by an i.m., s.c., or i.d. route with Freund's adjuvant. Askonas and Humphrey (1958) showed a changed pattern of lymphoid tissues involved in Ab synthesis after i.m. primed rabbits were boosted by the i.v. route. The spleen and bronchial lymph nodes of the rabbits became active in Ab synthesis, whereas they had not been involved after the primary immunization. Presumably the changed pattern reflects new sites of activity established through recruitment of circulating memory cells to lymphoid tissue with recently acquired Ag. The B-memory-cellTH-memory-cell interactions appear to take place in the follicular marginal zones ,but new germinal centers may also be established. For booster immunizations with particulate Ags such as virions, bacteria, or cells (where danger of anaphylaxis is low), the i.v. route has merit, because distribution of Ag is broad and capture by lymphoid tissues is high (Herbert, 1978a). For soluble Ags (where danger of anaphylaxis may be high), an s.c or i.m route may be preferred. Herbert (1978a) has suggested that booster inoculations with emulsion adjuvants may not be optimal, as most of the Ag is unavailable. However, the most frequently used protocol for soluble protein Ags involves priming with Ag in FCA followed by a booster immunization of Ag with Freund's incomplete adjuvant. The elevation of serum titer achieved through booster immunization may be more prolonged when a depot adjuvant is used. Booster immunizations by the i.d. route may not be as efficient as other routes, because distribution of Ag may be very localized.
As noted above, i.v. or even i.p. booster immunizations with aqueous soluble Ags may result in systemic anaphylaxis, caused by rapid release of histamine and other preformed mediators from mast cells and basophils. Guinea pigs and rats are more susceptible than some other species, but anaphylactic shock can be induced in most species if classes or subclasses of Abs with capacity to sensitize mast cells and basophils have been produced during the primary response. Measures to minimize risk of severe anaphylaxis at the time of the booster include: (1) administration of the booster by the s.c. or i.m. route to achieve slow Ag delivery, use of a depot-forming adjuvant to achieve slow Ag release, or use of both of these methods; (2) desensitization of the animal; (3) immunization in a manner to avoid production of sensitizing Abs during the primary response; and (4) pre-dosing the animal with an anti-histamine before administering the booster. One may desensitize an animal with a small dose of aqueous Ag by the s.c. route (or even by the i.v. route) a few hours before the bulk of the Ag dose is given by the i.v. route. One may minimize production of sensitizing Abs during primary immunization by using an adjuvant (such as FCA) that biases the primary T-cell response to a TH1 response. In mice and presumably other species, TH1 cells secrete cytokines that bias the secondary Ab response to non-sensitizing classes and subclasses of Ab. Alum, on the other hand, biases the T-cell response to a TH2 response and production of sensitizing Abs (such as IgG1 and IgE in the guinea pig).
Other comments on route. It has been suggested that differences in the quality of the immune response may result from the choice of route, particularly as an effect of the initial immunization. For example, in rabbits the primary immune response to an Ag/FCA emulsion administered by the i.d. route appeared to be strongly biased to the delayed-type hypersensitivity reaction of cell-mediated immunity with only a transient low-titer serum Ab response. However, subsequent aqueous Ag administered i.v. induced a high-titer secondary serum Ab response (Leskowitz and Waksman, 1960). Boosters with Freund's incomplete adjuvant s.c. or i.m. also seem to induce a strong humoral immune response after Ag/FCA priming. This information can be interpreted in terms of the TH1/TH2 model of T-helper cell responses, such that the primary immunization resulted in an initial TH1 response and the booster caused conversion to a balanced TH1/TH2 response. However, it is difficult to assess the individual contributions of adjuvant and route. Adjuvants clearly affect the type of T-helper cell response. Additionally, density of Ag on the Ag-presenting cell may also influence the type of T-helper cell response; a high density appears to bias the response towards TH1, whereas a low density appears to bias the response to a TH2 response (Janeway and Travers, 1994). The cytokines in the microenvironment during activation of the T-helper cells appear to be the deciding factors; as such, local tissue reactions and the nature or activation state of the Ag-presenting cells in the injection site (infection site) and draining lymph nodes may all contribute to the mix of cytokines in the microenvironment.
Chickens and other poultry deserve special comment concerning injection routes and other aspects of the immunization protocol, because avian species differ from mammals in the distribution and character of their lymphoid tissues. Development and diversification of their B-cell repertoire differs from that of mammals and is confined to an early development period (Reynaud et al., 1994). Nevertheless, chickens have an extensive repertoire and they develop a form of B-cell memory (Seto, 1980). The IgY found in plasma and in egg yolk (called IgY for yolk), differs significantly in structure from mammalian IgG (Parvari et al., 1988). It has properties in common with both IgG and IgE of mammals and may be representative of a "common ancestor" before divergence of IgG and IgE in mammals. Chickens and other avian species have abundant lymphoid tissues associated with their gastrointestinal tract, and have well-developed Harderian glands in the occipital orbits, but they lack the extensive network of peripheral lymph nodes common to mammals (Pink, 1992). For chickens, Glick (1986) has described small lymphoid accumulations along the posterior tibio-popliteal and lower femoral veins that behave as true lymph nodes, in that they have both afferent and efferent drainage and they respond to injection of Ag in the footpads with formation of germinal centers and generation of plasma cells. Ag injected in the lower leg muscle may be entrapped in these lymphoid tissues, but the majority of Ag injected into the pectoral muscle mass probably accumulates in the primary and accessory spleens. Many investigators inject the Ag with an oil-emulsion adjuvant in the breast muscle or leg muscle. For some Ags, Freund's incomplete adjuvant may be as effective as FCA (Shimizu et al., 1989). Immunostimulators that are effective in the chicken are different from those of many mammals. Ribi's adjuvant formulation designated for chickens uses a Salmonella typhimurium mitogen in place of the cell wall components used in adjuvant formulations for mammals.
Immunization Schedules
Schedules for booster immunizations are dependent on the response of the animal after the primary immunization, which in turn reflects the Ag quantity and quality, adjuvant, and injection route (or routes) (Cooper et al., 1991; Herbert, 1968; Herbert and Kristensen, 1986; Hu and Kitagawa 1990; Siskind et al., 1968). When an animal that has responded maximally is given a booster dose of Ag too soon, the immune response may be suppressed rather than enhanced. However, when an animal that has responded less than maximally is given a booster dose at 3 to 6 weeks after the first Ag dose, enhancement of the serum Ab titer usually occurs. Ideally, one follows the serum Ab titer in an immunized animal and gives a booster injection of Ag only after the Ab titer has plateaued or begun to decline. In practice, the investigator who "simply" wants to make an Ab to his/her favorite protein follows a preset schedule. In this case, patience is a virtue when selecting the time interval between the primary immunization and the booster immunization. One interpretation of why Ab response is improved with an interval of a few weeks between priming and boosting is that the memory B cells and memory T cells have been dispersed to additional lymph nodes resulting in potential involvement of more lymphoid tissue in the secondary response. This dispersal, not yet evident at day 8 post-priming (Stavitsky and Folds, 1972), probably begins during the late stage of the germinal center reaction (perhaps around day 10 of the primary response), which is biased towards production of memory B cells (Burton et al., 1994). Another interpretation is that excess Ag administered early in the response will induce suppression of the TH cells.
Herbert (1968), working with mice, showed the effect of ovalbumin booster immunizations (s.c.route, Ag in saline) at various times after a priming dose of 2 mg ovalbumin (a very high dose) in Freund's incomplete adjuvant (s.c. route). A booster at 20 days post priming gave a transient increase in serum Ab titer with a subsequent long-term depression, whereas a booster at 40 days (about the time the serum Ab titer plateaued) or later resulted in a prolonged increase in Ab titer.
Hu and Kitagawa (1990) and Hu et al. (1989b; 1990a) studied effects on serum anti-hapten Ab of various booster immunization schedules when mice were primed with a high dose (400 Tg), optimal dose (10 Tg), or low dose (3 Tg) of a hapten-carrier conjugate. In the study with high-dose priming (Ag administered i.p. in FCA), multiple or single booster immunizations were given over a period of ~6 weeks. A single booster at ~5 weeks resulted in a greater serum anti-hapten Ab response than four booster doses (200 Tg conjugate in FIA). When working with "optimal" Ag doses, the interval (2, 4, 6, or 8 weeks) between priming and a single booster (10 Tg in FIA) seemed less important than the number of boosters. The greatest enhancement of serum Ab titer was achieved with a single booster, whereas an increased number of boosters, administered at 2 week intervals, progressively diminished the effect of the booster. When the mice were primed with a suboptimal dose of Ag (3 Tg with alum as adjuvant) and boosted with a suboptimal dose (3 Tg), a greater number of boosters (1, 2, or 3, each with 3 Tg) administered at 2-week intervals increased the serum Ab titer. The interval (2, 4, or 6 weeks) between priming and a single booster immunization had a significant effect. When the booster was delayed, the Ab response was greater. A single booster at 6 weeks post priming resulted in ~600 Tg/mL hapten-specific IgG Ab in serum 2 weeks post booster, a value approximately 3- to 4-fold greater than when a single booster was administered at 4 weeks and IgG Ab was measured 2 or 4 weeks later. The 600 Tg/mL exceeded the value achieved with three boosters (~530 Tg/mL), although the difference was not statistically significant.
Some generalizations can be made about the immunization schedule. If the first and second immunizations are done without a depot-forming adjuvant, the Ab titer will usually peak within 1-2 weeks of the immunization, and subsequent booster injections may need to be administered frequently (monthly intervals) to maintain a high titer. If the primary immunization is done with a depot-forming adjuvant, titers may remain high for months, even with intermittant bleedings for harvesting the serum Ab. Intermittant bleeding of a hyperimmunized animal actually appears to help maintain a high serum Ab level, probably because it removes some of the suppressive effect of circulating Ab resulting in more Ab synthesis. This regulation probably occurs in the secondary follicles of spleen and lymph nodes where Ag-Ab complexes are stored on the FDCs. When the Ab titer drops, some of the Ab molecules complexed to Ag held by the FDCs will dissociate, thereby exposing free epitopes on the Ag. Memory B cells may then be stimulated by the Ag to become plasma cells, which will then secrete more Ab. Animals may also be rested for long intervals between boosting. Even when serum Ab titers have dropped to relatively low levels, a booster injection into an animal that had previously established a memory response will usually reestablish a high serum Ab titer.
Choice of Adjuvant
The large number of adjuvants available today reflects the lack of a single ideal adjuvant and the continuing search for substances with minimal toxicity and maximal immunostimulatory activity. It is probable that no single ideal adjuvant will be found for the many purposes investigators have in mind. The adjuvants used most frequently for Ab production in laboratory animals--Freund's adjuvants, Ribi Adjuvants®, and Hunter's TiterMax®--are discussed by Jennings (pp. 119-125 of this issue). The brief descriptions of other selected adjuvants, given below, should help acquaint investigators with some additional choices. Because Freund's adjuvants, still the most frequently used adjuvants in the United States, have the potential to cause significant pathology, investigators are often encouraged to consider alternatives.
When selecting an adjuvant, the investigator needs to consider the character of the protein Ag (such as the net charge of the protein Ag, its hydrophobic or hydrophilic properties, and its size) and the species chosen for immunization. An oil-in-water emulsion adjuvant may work better with hydrophobic or amphipathic (part hydrophobic, part hydrophilic) proteins than with very hydrophilic proteins. An acidic protein (net negative charge) may adsorb well to Al(OH)3 at neutral pH, whereas a basic protein will not. A small peptide, if not conjugated to an immunogenic carrier protein, may need to be encapsulated in or associated with liposomes or ISCOMs along with other peptides to serve as T-cell epitopes. The protein itself may need to be modified to make it compatible with a particular adjuvant. For example, a protein without any exposed hydrophobic character may need to be lipidated to enhance its association with the membrane of liposomes or the microdroplets of an oil-in-water emulsion. Charged groups may need to be added to proteins to enhance their association with alum. The investigator also needs to consider the fact that some adjuvants bias the secondary immune response towards expression of particular classes or IgG subclasses (Ig isotypes). Although guidelines for selecting the best adjuvant for a given Ag and given species are only beginning to be developed, much progress is being made. The next few years of adjuvant research should provide investigators with an improved set of guidelines.
Quantity of Adjuvant and Ratio of Ag to Adjuvant
Selection of the quantity of adjuvant is dependent primarily on the particular adjuvant and secondarily on the quantity of Ag to be used. Quantities suggested by manufacturer's have been derived empirically. Most adjuvants have an optimal protein:adjuvant ratio for optimal function. This can be understood at one level in terms of the need to deliver a threshold amount of protein to the Ag-processing/Ag-presenting cells in order for them to become efficient Ag-presenters. Perhaps a hundred or more peptide/MHC complexes are needed on the surface of an APC in order for it to form a stable conjugate with a T cell and deliver a threshhold signal. If the adjuvant vehicle, such as an oil droplet, that will be ingested by an APC contains only a few molecules of the protein Ag, insufficient peptide would be generated for the APC to carry out activation of a T cell. For alum as adjuvant, saturation of the Al(OH)3 with Ag appears to be counter-productive; open adsorptive sites apparently allow in vivo adsorption of Ig molecules to the alum particles and thereby help target the particle to relevant immune cells with Fc receptors for Ig molecules (Cooper, 1994).
SELECTED ADJUVANTS AND THEIR PROPERTIES
Montanide ISA Adjuvants® (Seppic, Paris, France)
These are a group of oil/surfactant-based adjuvants in which different surfactants are combined with either a non-metabolizable mineral oil, a metabolizable oil, or a mixture of the two. The variety of surfactant and oil combinations gives the investigator a choice. Various individual adjuvants among the Montanide ISA group of adjuvants are used as water-in-oil emulsions, oil-in-water emulsions, or water-in-oil-in-water emulsions; the different adjuvants accommodate different aqueous phase:oil phase ratios. The surfactant for Montanide ISA 50 or ISA 70 (ISA = Incomplete Seppic Adjuvant) is mannide oleate, a major component of the surfactant in Freund's adjuvants. The surfactants of the Montanide group undergo quality control to guard against contamination by substances that could cause excessive inflammation. The performance of ISA 50 (50:50::oil:aqueous Ag solution) or ISA 70 (70:30::oil:aqueous Ag solution) is said to be similar to Freund's incomplete adjuvant for Ab production, but the inflammatory response is frequently less. The manufacturer recommends preparation of water-in-oil emulsions such that microdroplets of the aqueous phase are 1 Tm or less and of homogeneous size.
Syntex Adjuvant Formulation (SAF)® (Chiron Corporation, Emeryville, California)
Allison and co-workers (Allison and Byars, 1991; Byars and Allison, 1995) developed SAF® as an alternative to Freund's adjuvant. It is a preformed oil-in-water emulsion stabilized by Tween 80 and the pluronic polyoxyethylene/polyoxypropylene block copolymer L121. The copolymer behaves as a surfactant and an adhesive for adsorbing protein Ags. SAF uses the metabolizable oil, squalene, and a low toxicity, high immunostimmulatory derivative of muramyl dipeptide, thr-MDP. SAF activates complement by the alternative pathway, thus helping to activate other cells in the immune system and to target the Ag to the follicular dendritic cells of the lymph nodes. It is said to bias the humoral immune response to IgG2a in the mouse (a TH1- supported response). However, as with other oil-in-water adjuvants, it works better with proteins that have some hydrophobic aspect to promote their adherence to the oil droplets. This adjuvant is said to bias the Ab response to native epitopes, as the protein is not subjected to harsh denaturing conditions when mixed with the adjuvant as a preformed emulsion.
Aluminum Salt Adjuvants
These adjuvants are used with protein Ags in two ways: (1) as alum-precipitated vaccines, and (2) as alum-adsorbed vaccines (Cooper, 1994; Harlow and Lane, 1988;Lindblad, 1995; Nicklas, 1992). Investigator-generated or commercially available Al(OH)3 (Alhydrogel® - Superfos of Denmark/Accurate Chemical & Scientific Co., Westbury, New York) can be used to adsorb anionic proteins in the ratio of 50- ~200 Tg protein/mg aluminum hydroxide. Adsorption of protein is dependent on the pI of the protein and the pH of the medium. A protein with a lower pI adsorbs to the positively charged aluminum ion of Al(OH)3 more strongly than a protein with a higher pI. Aluminum salts are generally weaker adjuvants than emulsion adjuvants, and because of their generally mild inflammatory reactions, safety, and efficacy for generating memory, they are the primary adjuvants used in humans. The are generally not effective with small peptide Ags. When used in larger quantity in laboratory animals than would be used in humans, the inflammatory reactions that may occur at the site of injection will generally resolve within a few weeks although chronic granulomas may occasionally form (White et al., 1955). The mineral adjuvants work by concentrating and aggregating the Ag, establishing a depot of Ag that is released slowly over a period of 2-3 weeks, and by activating macrophages. Activation of complement has been debated. They bias the immune response towards humoral immunity (a TH2 response in the TH1/TH2 model of cytokine regulation) with the major responsive isotypes being IgG1 and IgE in the mouse (Cooper, 1994; Kenney et al., 1989). The effectiveness of aluminum salt adjuvants has been increased in experimental studies by the addition of detergents, MDP, or Bordetella pertussis, but the inflammatory reaction is generally increased as well. Due to the short-term depot effect of alum, booster injections may be needed more frequently than with water-in-oil emulsions.
Algammulin®, a supension of microparticles (1-2 Tm) of gamma-inulin coated alum (10:1), has adjuvant properties with anionic protein Ags that exceed those of either of the components, alum or inulin (ooper, 1994; Cooper et al., 1991). Under optimal conditions of immunization of mice with KLH and Algammulin® or KLH and FCA as adjuvant, Cooper and co-workers found total serum IgG Ab concentrations to be similar for the two adjuvants. The anti-KLH IgG isotype profile with Algammulin® as adjuvant was balanced between IgG1, IgG2a, IgG2b and IgG3. Under conditions of low dose KLH for primary and secondary immunizations, the secondary response to the booster immunization was higher when Algammulin, rather than saline only, was used in the booster.
Nitrocellulose-adsorbed Protein
Nitrocellulose-adsorbed protein can be used for immunization without desorption of the protein in vitro, as desorption of protein from nitrocellulose (NC) paper will occur in vivo giving a desirable slow release of Ag over a period of 2 weeks to 2 months (Nilsson and Larsson, 1992). The nitrocellulose itself is essentially inert, causing little if any inflammatory response and no anti-NC Ab response. Either intrasplenic or subcutaneous deposition of the nitrocellulose paper with ng to Tg quantities of Ag has been successfully used for Ab production. Approximately 100 Tg of protein can bind to 1 cm2 of NC, so that only small quantities of NC need be introduced into the animal. The Ab response is not as vigorous as with FCA/Ag emulsions and may require one or more booster immunizations to be detected, but this method is particularly advantageous for situations in which only small quantities of pure Ag can be obtained as a band from an electroblot of a gel. Antibodies have even been raised to NC-adsorbed protein administered after the protein had been stained with Coomassie Blue, or after the NC membrane had been pelleted with dimethyl sulfoxide (Forrest and Ross, 1993). However, the investigator should be aware that Abs raised to the reduced or otherwise denatured protein may not react well with the native protein. On the other hand, such Abs may be desirable for use on electoblots (Western blots). Membranes other than nitrocellulose may also be used in a similar manner (Nilsson and Larsson, 1992).
Encapsulated or Entrapped Ags have been prepared in several ways that permit sustained slow release of Ag and, in some cases, release of immunostimulators as well. Some examples include liposome-entrapped or liposome-associated Ag (Gregoriadis et al., 1993, 1995); ethylene-vinyl acetate copolymer (EVAc)-entrapped Ag (Niemi et al., 1985); and poly (methylmethacrylate)-entrapped Ag (Kreuter, 1993). Additionally, pulverized slices of polyacrylamide gel containing an isolated protein (from a gel electrophoresis procedure) may be prepared for injection (Amero et al., 1988; Harlow and Lane, 1988). The complexity of preparation is a drawback for some of the encapsulated/entrapped Ag preparations, but in special situations the potential of an encapsulated/entrapped Ag to generate a significant immune response may make preparation of these materials worthwhile.
Use of acrylamide-entrapped Ags for immunization has been described by Amero et al. (1988). Protein bands of interest may be identified by staining the gel to be used for Ag preparation or by staining a parallel test strip and storing the remainder of the preparative gel in the frozen state. Some precautions are needed to assure that the injected preparation is free of toxic materials such as (1) glutaraldehyde from in-gel cross-linking of proteins, or (2) unpolymerized acrylamide, which is neurotoxic and immunotoxic (Zanelli et al., 1994). For buffer exchange and removal low molecular weight toxic materials, gels may be soaked for a short time (10 minutes x 3) in a physiological buffer without significant loss of proteins. Removal of excess SDS, an ionic detergent, may be accomplished by presoaking the gel slice in buffer containing 20% ethanol prior to physiologic buffer exchange. The wet gel may be pulverized by multiple passes through a 20 gauge needle, or the gel may be freeze-dried, pulverized, and resuspended in saline prior to injection. Such preparations have been used with Freund's adjuvants, but the acrylamide itself has some adjuvant properties.
Among the biodegradable, biocompatible polymers used for encapsulation, poly(DL-lactide-co-glycolide), a material used for sutures, demonstrates favorable characteristics for use in bulk-prepared vaccines. Construction and administration of different size microspheres in a single dose may accomplish timed release of Ag in a way that mimics primary and booster injections (Eldridge et al., 1991). However, the preparation is rather complex for use with the occasional Ags prepared for injection by individual investigators, particularly when the Ag is available in very limited quantity.
Liposomes, which are vesicles made of lipid bilayer membranes, may be used as vehicles for Ag delivery. Liposomes may be multilamellar or unilamellar; stability depends on size and the mixture of lipids, cholesterol, phospholipids, and sugars used in preparing the membranes. Liposomes with with entrapped Ags in a native state may be prepared in an investigator's laboratory by a method recently described by Gregoriadis (1995). This method, which involves dehydration and rehydration of small unilamellar vesicles to accomplish Ag entrapment, can yield small vesicles suitable for lymphatic delivery when injected i.p., i.m., or s.c. Alternatively, liposomes with membrane-associated Ags may be prepared (Gregoriadis, 1993). In addition to B-cell Ags, other substances such as immunostimulators (e.g., cytokines), peptides to serve as T-cell epitopes, or targeting molecules may be associated with the liposomes. Small vesicles will be rapidly absorbed whereas large multilamellar vesicles may form a temporary injection-site depot. Liposome-entrapment is a desirable strategy for preparation of venoms and toxins for injection in the course of Ab production.
ISCOMs or Immune-Stimulating COMplexes are Ag-modified saponin/cholesterol micelles that form stable cage-like structures, 30-40 nm in diameter Claassen and Osterhaus, 1992; Dalsgaard et al., 1995). The saponins (Campbell, 1995) of the Quil A group are low toxicity immunostimmulating substances that are responsible for the cage-like protective structure. Protein Ags having some hydrophobic aspect are incorporated into the cage in the presence of small quantities of phospholipid and detergent; and even proteins without a hydrophobic aspect can be altered by linking fatty acid molecules or by partial denaturation to permit the proteins to assemble in the cages. The ISCOM-associated Ag molecules do not form an Ag-depot at the site of injection (s.c. or i.m.), but are instead rapidly transported to draining lymph nodes (Morein et al., 1993). They may also be administered i.p., i.v., or even orally. Quantities of Ag as low as 1 Tg have elicited a significant immune response, consisting of both cell-mediated and humoral immune responses. ISCOMs can be successfully prepared in an investigator's laboratory (Coligan et al., 1995).
Gerbu® Adjuvant (Gerbu Biotechnik GmbH, Gaiberg Germany/C-C Biotech, Poway, California)
This is a new aqueous phase adjuvant that does not have a d