Advanced Drug Delivery Reviews 43 (2000) 2943

www.elsevier.com / locate / drugdeliv

Application of DNA vaccine technology to aquaculture

Joel Heppell a , Heather L. Davis a,b,c , *
a b

Loeb Health Research Institute at the Ottawa Hospital, 725 Parkdale Avenue, Ottawa, ON, Canada K1 Y 4 E9 School of Rehabilitation Sciences, Faculty of Health Sciences and Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, Ottawa, ON, Canada c CpG ImmunoPharmaceuticals Inc., Wellesley, MA, USA

Abstract The aquaculture industry needs to augment its global production and efciency to meet the increasing consumer needs for sh and shellsh products. Unfortunately, infectious diseases have been a major impediment to the development and protability of sh farms. While vaccines offer the most efcient way to control infectious pathogens, current products have only been successful against some diseases. These are mostly bacterial, and there are still several important diseases, mainly of viral and parasitic origin, for which no prophylactic treatment exists. DNA vaccines, compared to traditional antigen vaccines, have several practical and immunological advantages that make them very attractive for the aquaculture industry. The early success of DNA vaccines in animal models was very encouraging, but sh are unique in many aspects, and ndings with other classes of vertebrate, namely mammals and birds, do not necessarily apply to aquatic animals. However, more recent studies with reporter genes showed that sh cells efciently express foreign proteins encoded by eukaryotic expression vectors. A piscine-specic backbone vector might eventually improve immune responses to DNA vaccines, but there is already strong direct evidence for the induction of protective immunity with currently available plasmids. Immune responses to plasmid DNA injected intramuscularly (IM) into sh are characterized by the production of antibodies, which have been shown to be neutralizing in two different viral disease models. There is also indirect evidence suggesting the induction of cell-mediated immunity. Despite this evidence, immune responses to DNA vaccines have only been poorly characterized in sh because of the limited knowledge of the piscine immune system, and the small number of studies on the subject. Apart from optimizing the efciency of DNA vaccines, other important issues, such as safety and production cost will be determinants for the potential application of this technology in commercial sh farms. Alternative methods of administration will also have to be developed for small sh and low-valued species, for which IM injection is not practical and / or cost effective. 2000 Elsevier Science B.V. All rights reserved.
Keywords: Nucleic acid vaccine; Fish; Aquaculture; Disease control; CpG motifs

Contents 1. Introduction ............................................................................................................................................................................ 1.1. The need for sh vaccines ................................................................................................................................................ 1.2. Advantages of DNA vaccines ........................................................................................................................................... 30 30 31

*Corresponding author. Tel.: 1 1-613-798-5555 ext. 7682; fax: 1 1-613-761-5354. E-mail address: hdavis@lri.ca (H.L. Davis). 0169-409X / 00 / $ see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0169-409X( 00 )00075-2

30

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43 31 31 31 32 32 33 33 34 34 34 35 35 36 36 37 38 38 38 38 39 39 40 40

2. Plasmid constructs................................................................................................................................................................... 2.1. Control elements .............................................................................................................................................................. 2.2. Antigen encoding gene ..................................................................................................................................................... 2.3. DNA as an immunostimulatory molecule for sh ............................................................................................................... 3. Methods of administration ....................................................................................................................................................... 3.1. Injection .......................................................................................................................................................................... 3.1.1. Site of injection...................................................................................................................................................... 3.1.2. Dose and volume ................................................................................................................................................... 3.1.3. Location of foreign gene expression ........................................................................................................................ 3.1.4. Duration of expression and effect on sh tissues....................................................................................................... 3.1.5. Alternatives to needle injection ............................................................................................................................... 3.2. Oral and immersion vaccination ........................................................................................................................................ 4. Immune responses to DNA vaccines ......................................................................................................................................... 4.1. Antibody responses .......................................................................................................................................................... 4.2. Cellular immune responses ............................................................................................................................................... 4.3. Protection against live challenge ....................................................................................................................................... 5. Safety of DNA vaccines for sh ............................................................................................................................................... 5.1. Fate of injected DNA in sh ............................................................................................................................................. 5.2. Safety for the consumer .................................................................................................................................................... 5.3. Risks for other sh and the environment ............................................................................................................................ 6. Perspectives for DNA vaccines in the aquaculture industry......................................................................................................... 7. Conclusions ............................................................................................................................................................................ References ..................................................................................................................................................................................

1. Introduction The need for additional or alternate sources of sh and shellsh products has made the aquaculture industry the fastest growing segment of agriculture in the world [1,2]. The increased demand for aquatic animals, which is partly due to greater health awareness of consumers, and the decline or stagnation of natural harvests, have largely contributed to this rapid growth [1]. However, despite the signicant progress toward higher standards of productivity and quality that sh farming has made in developed countries, the industry has still not reached levels comparable to those found with other farmed animals such as swine and poultry.

1.1. The need for sh vaccines

Among problems that the sh and shellsh industries have to address, infectious pathogens are the most important [1,3,4]. It has been estimated that ten percent of all cultured aquatic animals are lost because of infectious diseases alone [5]. Outbreaks can cause severe losses to sh farmers and in many

instances, there are no prophylactic or therapeutic measures available. Chemicals and antibiotics can be used to control bacterial and parasitic diseases but these products often have undesirable side effects such as accumulation in the esh of animals, development of drugresistant strains, and contamination of the aquatic environment [6,7]. For viral infections, where there are no treatments available, the appearance of disease in facilities usually requires destruction of infected stock before starting anew. Consequently, most research efforts are focusing on prevention rather than treatment with chemicals, which should only be considered when vaccines are unavailable for specic pathogens or when they fail to prevent an outbreak. Vaccines have long proven their efcacy in herd protection, but in the aquaculture industry, they are in a relatively early phase of development. Most of the commercially available vaccines protect sh against bacterial diseases, and are simply made of inactivated bacteria applied by either immersion or injection with an oil adjuvant [3,8,9]. There are currently no vaccines against parasitic pathogens, and very few against viruses [9,10]. The high cost of

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

31

new product development combined with the relatively small size of the industry and low value of individual animals have largely contributed to this situation. In fact, many vaccines have been tested under laboratory conditions, but they were not commercially viable because of their prohibitive cost of production, insufcient protection, or their lack of safety [5,6,8,11,12].

there is no indication that it will be very different from what has been observed with other animals.

2. Plasmid constructs There are still very few reports on the use of DNA vaccines in sh, and except for research on transgenic animals, which is not the subject of this review, the question of designing a sh specic vector has never been addressed. Nevertheless, the major considerations in vector design for the purpose of DNA vaccines are presented in this section. Features not specic to sh, such as bacterial origin of replication and selection marker, have been reviewed elsewhere [23,24] and are not discussed herein.

1.2. Advantages of DNA vaccines

The rst experiments showing an immune response to plasmid-encoded antigens of infectious pathogens were published several years ago [1318]. This very promising technology immediately caught the widespread attention of scientists working in the eld of vaccine development. An early report in 1991 by Hansen et al. [19] showed expression of a reporter gene in carp muscle after injection of plasmid DNA, but it was several years before DNA vaccines were demonstrated in sh [2022]. For aquatic organisms, as for other farmed animals, DNA vaccines offer several advantages over classical antigen vaccines (i.e., live attenuated, whole killed and subunit vaccines). From a practical point of view, they are relatively inexpensive and easy to produce, and all DNA vaccines require identical production process. Multivalent vaccines can also be easily prepared by mixing together different plasmids, or including more than one antigen-encoding gene in a single vector for colinear expression, which will further reduce the cost of production. In addition, DNA is a very stable molecule, and does not need to be maintained in a cold environment during shipment or storage. Finally, the ease of cloning allows vaccines to be rapidly modied if needed. All these factors contribute to make DNA vaccines very attractive to sh vaccine manufacturers. DNA-based immunization also has immunological advantages over traditional methods of vaccination. As shown with mammals, they can induce strong and long-lasting humoral and cell-dependent immune responses without boost, similar to that conferred by live vaccines, but without the risk of inadvertent infection [23]. In sh, much less is known about the immune responses following plasmid injection, but

2.1. Control elements

Plasmid constructs used for expression of foreign genes in sh usually contain a mammalian expression vector backbone. Sequences for transcriptional control in the standard vectors (promoter, enhancer, intron, polyadenylation signal, etc.) seem to work efciently in aquatic animals. Various promoters have been assessed for their ability to drive expression of foreign genes in sh tissues (Table 1). Their efcacy varies considerably, but overall, the immediate early promoter of the cytomegalovirus (CMV) gives the best results. It is also the most widely used in DNA vaccines reported to date, and its potency has been demonstrated in sh vaccination trials [2022,33]. Other transcriptional elements in DNA vaccines have not been evaluated in sh, but there are indications from in vitro studies that mammalian introns might not always be correctly processed in sh cells [34]. Optional features of backbone vectors such as unrelated Th epitopes to provide additional T help, CpG immunostimulatory motifs, or genes for colinear expression of cytokines or costimulatory molecules, could eventually be added to DNA vaccines [23]. Although a better characterization of the sh immune system is required before similar sequences are incorporated into vaccines for aquaculture, a preliminary study suggests that mouse granulocytemacrophage colony-stimulating factor

32

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

Table 1 List of promoters tested in sh tissues. Promoter Glucocorticoid-responsive mouse mammary tumor virus (MMTV) Cytomegalovirus (CMV) immediate early (alone or with translational promoter) Fish species Rainbow trout (Oncorhynchus mykiss) Rainbow trout (Oncorhynchus mykiss) Zebra sh (Brachydanio rerio) Goldsh (Carassius auratus) Xiphophorus sp. Rainbow trout (Oncorhynchus mykiss) Tilapia (Oreochromis niloticus) Common carp (Cyprinus carpio) Common carp (Cyprinus carpio) Common carp (Cyprinus carpio) Common carp (Cyprinus carpio) Rainbow trout (Oncorhynchus mykiss) Xiphophorus sp. Xiphophorus sp. References [25] [21,22,25,26] [21] [2729] [30] [25] [31] [19] [19] [19] [19] [32] [30] [30]

Carp b -actin SV40 early Rabbit b -cardiac myosin heavy chain (MHC) Human MxA Articial (based on human MxA) Herpes simplex virus thymidine kinase (tk) with the CMV enhancer with the SV40 enhancer

(GMCSF) gene coinjected with an antigen-encoding gene could modulate cell-mediated immune responses in goldsh [35].

2.2. Antigen encoding gene

In theory, any gene coding for a protein of the pathogen that induces a protective immune response in the host species can be used in DNA vaccines, as long as the codon usage of the gene allows expression in sh cells. A problem with codon usage could eventually arise with genes from some bacteria or lower parasites, but it is unlikely to happen with viral genes that are normally expressed in host cells using cellular machinery. To induce humoral immune responses, B-cells must rst meet circulating antigen to be activated. Thus, secretion of the foreign protein can inuence the antibody production. However, humoral responses are possible even when the antigen is not secreted. For example, antibodies against the cottontail rabbit papilloma virus (CRPV) major capsid protein (L1) were detected after DNA immunization, even when L1 had a nuclear localization signal [36]. In the case of non-secreted antigen, B-cells may not be fully activated unless antigen is released from transfected cells, such as following lysis by antigenspecic cytotoxic T-lymphocyte (CTL). It has been

shown in mice injected with hepatitis B surface antigen (HBsAg), that appearance of antigen-specic antibodies is delayed by a few weeks if the protein is not secreted [37]. Consequently DNA vaccines may include a signal peptide to direct secretion of the encoded antigen, but this does not appear to be an absolute requirement.

2.3. DNA as an immunostimulatory molecule for sh

Strong immune responses induced by DNA immunization are not dependent solely on expression of antigen. It is now well established that the DNA molecule itself can act as an adjuvant to enhance the immune response in mammals [38,39]. In fact, it has been shown that short DNA sequences containing unmethylated CpG dinucleotides in specic base contexts can directly stimulate immunocompetent cells such as monocytes, B-cells, macrophages and dendritic cells; these immune stimulatory motifs are called CpG-S motifs [39]. Adversely, other CpG motifs that neutralize or block immune activation by CpG-S motifs are called CpG-N motifs [40,41]. Very little is known on the effect of CpG motifs in sh, but preliminary experiments suggest that stimulation can occur as in mammals, although the optimal CpG-S motifs for sh are probably different from

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

33

those previously determined for mice [35,42]. Plasmid DNA vectors used to vaccinate sh certainly contain CpG-S and CpG-N motifs in their sequences. It is possible that a CpG optimized backbone vector, containing an optimal number of sh-specic CpG-S motifs and a reduced number of CpG-N motifs could improve the immune response in sh, and possibly allow administration of smaller doses of DNA.

no information is available on the effects of different routes of administration on immune responses, owing to the limited studies and the relatively poor characterization of the piscine immune system.

3. Methods of administration Vaccine administration to aquatic animals poses obvious technical problems not encountered with other animals. Nevertheless, in developed countries, large-scale vaccination is prevalent for high valued species such as salmon [11,43]. Methods used to immunize sh can be divided into three major categories, with several variations in each: injection, immersion, and oral delivery [7,8,44].

3.1. Injection 3.1.1. Site of injection Although stressful for the animals and labor intensive, injection vaccination is usually the most immunologically effective method to immunize sh [7,8]. For DNA vaccines, it is thus far the only method that has been reported in vaccination trials. Puried plasmid DNA in a small volume of saline or buffer is usually injected intramuscularly in the ank, below or close to the dorsal n, but other sites have also been tested (Table 2). Injection in the gills does not result in detectable levels of reporter gene activity [32], while the outcome with the peritoneal route was shown to be variable [28,32]. In mice, the site and method of injection of DNA vaccines have a strong inuence on the type of immune responses elicited [23]. For example, intramuscular (IM) needle injection induces a predominantly Th1 type response, with a high IgG2a:IgG1 ratio, interferon (IFN)-g production and low level of interleukin (IL)-4. On the other hand, epidermal gene gun inoculation commonly induces a Th2 response, characterized by the predominance of IgG1 over IgG2a, less IFN-g and higher IL-4. Intradermal (ID) injection of DNA with a needle can induce either Th1 or Th2 proles [45]. In sh, almost

3.1.2. Dose and volume The effect of dose and volume of the injected DNA has been studied more rigorously (Table 2). Larger volumes are usually associated with a reduction of the reporter gene activity, but less variation between individual sh. Some authors have also noted that an excessively high dose of DNA can result in a lower expression level of the foreign gene [30,32]. Typical doses for sh fall in the range of 150 mg of DNA, in a volume of 1050 ml. However, much lower doses have been found to be effective [21]. The optimal dose of DNA probably varies according to the species and size of the animal, but it does not increase proportionally to the weight of the sh. Moreover, it has been reported that the physiological condition of the animal also inuences the expression level. Young and fast growing carp showed much higher levels of chloramphenicol acetyl transferase (CAT) reporter gene activity than older sh [19]. Interestingly, luciferase reporter gene expression driven by the CMV promoter is higher in sh than mouse muscle for the same dose of DNA [21]. It is not known whether this is due to an increased stability of the luciferase protein in sh compared to mammalian cells, to a higher expression level, or to a better transfection efciency. However, it suggests that low doses of DNA, compared to those injected in mammals, could be used to vaccinate sh. In fact, doses as low as 0.1 mg of DNA per sh have been reported to be as effective as 10 mg for inducing protective immunity [33]. This bodes well for the potential application of DNA vaccines in sh farms. 3.1.3. Location of foreign gene expression There are apparent discrepancies in the literature regarding the site of foreign gene expression after IM injection of plasmid vectors. In large sh, expression of reporter genes seems to be mostly restricted to the site of injection, that is in myocytes as well as in cells inltrating the muscle tissue or epithelial cells lining small capillaries [26]. Transfection of nonmuscle cells, especially professional antigen-present-

34

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

Table 2 Methods of injection and doses of DNA tested in sh tissues. Method a IM single injection Site Flank
b

Gene d CAT CAT Luciferase, IHNV-G and -N lacZ, luciferase lacZ, luciferase VHSV-G and -N VHSV-G and -N lacZ lacZ lacZ, luciferase VHSV-G, IHNV-G and rabies virus G luciferase lacZ luciferase lacZ

Dose (mg) 12.5, 25, 50 and 100 5, 10 and 50 10, 25 and 50 10, 25, 50 and 100 0.01, 0.1, 1, 10 and 50 5, 10 and 50 0.1, 1 and 10 35 1, 5, 25, 50 and 125 5, 10, 20 and 40 30

Volume (ml) 100 0.1 100 and 200 100 5, 10, 25, 50 and 100 25 25 50 50 5 and 6 100

References [19] [31] [20,25] [32] [21] [22] [33] [27] [28,29] [30] [26]

IM multipoint injection IP injection Gill injection ID multipoint injection Particle bombardment

Flank

Peritoneum

25 50 25 50

100 50 100 50

[32] [28] [32] [28]

Gills Ventral c

Flank Eye Flank (midsection) Flank (close to tail)

Luciferase Luciferase Luciferase Luciferase

0.3, 2, 5, 11 and 25 11 11 11

N.A.e N.A. N.A. N.A.

[32] [32] [32] [32]

Abbreviations: IM 5 intramuscular, IP 5 intraperitoneum, ID 5 intradermal. All injection sites located below or rostroventral to the dorsal n and above the lateral line are included in this category. The exact location of the injection varies between experiments. c Midline between the pectoral and anal ns. d Genes: CAT 5 chloramphenicol acetyltransferase; IHNV-G and -N 5 infectious hematopoietic necrosis virus G and N genes, respectively; VHSV-G and -N 5 viral haemorrhagic septicaemia virus G and N genes, respectively. e Not applicable.
b

ing cells, could contribute to improved immune responses. In small sh, expression of the luciferase reporter gene injected IM was detected in different organs, including gills, spleen and kidney, but was highest in muscles [21,25]. Differences between large and small sh are most likely due to more rapid spread of injected DNA from the injected muscles in small sh, but there may also have been variations in the protocol, sensitivity of the assay and animal model used. Although promoters can also inuence the location of foreign gene expression, the same CMV promoter was used in these studies.

3.1.4. Duration of expression and effect on sh tissues IM injection induces strong and long-lasting expression levels of reporter genes, although different times of peak activity and total duration of expression have been reported (Table 3). With the luciferase gene, Anderson and coworkers [25] observed a diminution of enzyme activity after 7 days, but this was not conrmed by others [21,32]. In mice, luciferase transfected cells are not destroyed by CTL because no (or very low) immune response is raised against the luciferase protein due to its poor im-

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43 Table 3 Duration and peak activity of reporter genes injected into sh muscle. Reporter gene Luciferase Fish species Rainbow trout (Oncorhynchus mykiss) Peak activity (days) 7 60 2.5 2.5 6 21 2 Duration (days) . 115 . 60 . 84 . 112 . 10 . 70 ,7

35

References [25] [32] [21] [21] [30] [28] [31]

lacZ Chloramphenicol acetyl transferase (CAT)

Zebra sh (Brachydanio rerio) Xiphophorus sp. Goldsh (Carassius auratus) Tilapia (Oreochromis niloticus)

munogenicity [46] and, expression has been detected for nearly 2 years [47]. In any event, it appears that expression of an antigenic protein will eventually be curtailed by cellular immune responses against the antigen-expressing cells. In mice, most muscle cells expressing immunogenic proteins are destroyed between day 10 and 20 following gene delivery [46]. In sh, this elimination process also seems to occur, but more slowly. The number of b-galactosidase expressing cells in sh muscle was found to decrease slowly between day 21 and 70 after injection of the lacZ gene alone [29], but coinjection with the mouse GMCSF gene was shown to have a marked effect on the rate of disappearance of transfected muscle bers [35]. With the G gene of VHSV, there is indirect evidence that the number of cells expressing the viral protein at day 84 is reduced by 90% compared to the maximum number that is found between 2.5 and 14 days [21]. Histological studies with the lacZ reporter gene showed that injection of pure DNA solution into rainbow trout muscle does not cause permanent tissue damage. The presence of inammatory cells in the needle track has been observed early after injection [32], but there were no signs of necrosis, and tissue appeared completely regenerated without scarring by 28 days following injection [21]. In addition, other negative side effects such as visceral adhesion and grown retardation, observed with oiladjuvanted antigen vaccines [48], have never been reported with DNA administration.

3.1.5. Alternatives to needle injection As an alternative to direct injection with a needle, gene delivery by particle bombardment has been assessed in sh [32], however, results were not as good as with direct needle injection, both for expression level and proportion of animals expressing the reporter gene. Even if the efciency of the technique is improved, it is unlikely that this technology will be widely employed on sh farms because of its higher cost and practical limitations (e.g., the DNA falls off the gold particles in a humid environment). 3.2. Oral and immersion vaccination
For sh farmers, oral vaccines are the ideal method of immunization because they are easy and inexpensive to administer, and they are not stressful for the animals. There are also immunological advantages associated with oral delivery, such as the induction of mucosal immunity. Very little information is available on mucosal immunity in sh, but it has been shown to be stimulated by antigens [49]. Oral delivery also has some practical and immunological drawbacks. First, it is not possible to determine the dose of vaccine ingested by each animal, although this is not that important as long as good herd immunity is induced. More importantly, at least with antigen vaccines, oral delivery appears to be less effective than injection or immersion for vaccinating sh [43]. Immersion vaccination, where animals are placed

36

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

in or sprayed with a concentrated vaccine solution, is more widely used than oral administration for antigen vaccine delivery, and normally provides better protection, but still not as good as injection. Compared to injection however, immersion vaccination requires less manipulation of the sh and can be used even for very small sh, but it does require more vaccine solution, especially for large sh. The greater effectiveness of immersion over oral delivery is thought to be due to the better absorption of antigen through the skin and / or gills compared to the gut, as well as degradation of the vaccine in the digestive tract [43]. Recent quantitative experiments have shown that both skin and gills participate in uptake of vaccine with immersion, with the skin playing the more dominant role [50,51]. However, this could depend on the vaccine since other studies have reported that oral ingestion is the principal route of antigen entry in bath-immunized sh [52]. At the time of writing this review, there were no reports of DNA vaccines administered to sh by oral or immersion routes. Using reporter genes, we have never observed detectable levels of gene expression following immersion or oral delivery of puried plasmid DNA to small sh. However, association of DNA with transfection agents such as liposomes or microcarriers can result in transfection of immersed sh [Heppell et al., unpublished results]. This suggests that immersion and oral vaccination of sh with formulated DNA vaccines might be feasible and effective.

cellular and humoral, that rely heavily on T- and B-lymphocytes respectively. As for mammals, specic humoral immunity in sh involves recognition and binding of circulating soluble antigens by Bcells, which then respond by producing and secreting antigen-specic antibodies. On the other hand, specic cell-mediated immunity involves T-cells recognition of antigens presented on cell surfaces in association with molecules of the major histocompatibility complex (MHC), which triggers induction of cytotoxic T lymphocytes (CTL) and T-cell secreting cytokines. In mammals, CD4 and CD8 T-cells recognize antigen presented by MHC II and MHC I molecules respectively, but it is not clear if this is also true in sh as there are no reagents to identify CD4- and CD8-like molecules [55]. DNA vaccines are known to trigger both nonspecic (via the CpG-S motifs) and specic immune mechanisms (via the expression of antigen). In the latter case, they efciently stimulate both humoral and cellular immune responses. As such, they behave more like a live vaccine, except without the risk of infection, and perform better than inactivated vaccines or puried macromolecules, which induce mainly humoral responses.

4.1. Antibody responses

DNA vaccines allow expression of antigens in situ. Therefore, foreign proteins, especially those of viral origin, can be presented to the immune system in their proper conformation with post-translational modications and intracellular trafcking identical to those of antigens produced during natural infection. Since many B-cell epitopes are conformational, presentation of antigenically relevant epitopes to the immune system may be more readily attainable with DNA vaccines than with other type of vaccines, such as inactivated or subunit vaccines, where conformational epitopes might be lost [24]. In mammals, the strength and longevity of antibody responses induced by DNA vaccination depend on the animal and antigen model. Humoral responses are usually long-lived in mice [56], but appear to be shorter in non-human primates [24]. A single dose of antigen-encoding DNA is sufcient to induce a protective immune response in some cases, but depending on the antigen, the route of administration

4. Immune responses to DNA vaccines The rst line of defense against infection for sh, as for other animals, is the nonspecic immune system. This includes physical barriers such as mucosal surfaces and skin, but also a variety of leukocytes (e.g., monocytes / macrophages, granulocytes and nonspecic cytotoxic cells) [53] and substances (e.g., lysozyme, complement, interferon, C-reactive protein, transferrin and lectin) that nonspecically inhibit the growth of infectious microorganisms [54]. The innate immune system is of major importance in sh. The second line of defense is the specic or acquired immunity. This can be divided in two arms,

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

37

and the animal species, boosts might be required [23]. The antibody isotype induced is generally IgG, although serum IgM and IgA have also been reported [57,58]. Few studies have examined the antibody response in sh following DNA administration. Comparison between published experiments is difcult because species and conditions used were different, but they all showed that DNA vaccines induce antigen-specic immunoglobulins in sh. Antibodies against b-galactosidase have been detected in goldsh as early as 7 days after injection of lacZ encoding DNA [29]. In rainbow trout, antibodies to the G protein of viral haemorrhagic septicaemia virus (VHSV) were detected 23 days after injection with a plasmid encoding the G gene [26]. The concentration of serum antibodies peaked at 3 to 8 weeks and remained high for several weeks after [20,29], although the number of antibody-forming cells appeared to decline rapidly [29]. In the rst report of a sh DNA vaccine, seroconversion was detected in only a small percentage of immunized rainbow trout (2 / 15) [20], but later studies demonstrated that it is possible for all vaccinated sh to have signicant levels of antibody [2628]. In fact, Russell et al. [27] showed that after IM injection of a DNA vector coding for b-galactosidase, goldsh seroconvert even more readily than mice. This may relate to the less developed connective tissue compartmentalization of muscles in sh compared to higher vertebrates, which could make piscine muscle bers more accessible to direct DNA transfer and increasing the possibility for DNA to spread to other tissues [21,27]. As in mammals, the antibody response in sh is dose-dependent, and seroconversion is incomplete when low doses of DNA are used [28]. Boosting on the third week following priming, does not seem to produce a stronger or earlier humoral immune response [20]. As well, Lorenzen and coworkers [22] have shown that the strength of the neutralizing activity in sera collected from rainbow trout immunized with plasmid DNA encoding the G gene of VHSV was not signicantly different before and after challenge with live virus. These results suggest that near maximal humoral responses are induced by a single dose of DNA vaccine. However, the effect of boost on the duration of protection has not been evaluated.

Only tetrameric forms of antibodies (IgM-like) have been originally isolated from sh serum, but it is now well established that sh have different antibody isotypes, although the functional relevance of this structural diversity is not quite clear, and considerable variations exist between piscine classes [59]. The type of immunoglobulin associated with IM injection of DNA in sh has not been determined. However, the avidity of the antibodies appears to be similar with DNA- and protein-immunization [28].

4.2. Cellular immune responses

The in vivo synthesis of antigen with DNA-based vaccination is a desirable feature because it mimics a natural infection by intracellular pathogens with respect to MHC class I antigen presentation. This is associated with strong cell-mediated responses including CTL. Such Th1-biased responses may be reinforced by the presence of CpG-S motifs in the plasmid backbone [60]. MHC class I-restricted, CD8 1 CTL have been demonstrated in lymph nodes or spleen of mice injected IM with plasmid DNA [24]. Generation of memory T lymphocyte responses has also been demonstrated with a variety of plasmid constructs [24] Cellular immune responses in sh are more difcult to assess owing to the relatively poor understanding of the piscine immune system and the general unavailability of inbred strains of animals and appropriate reagents. Nevertheless, there is indirect evidence for activation of this arm of the immune system in sh. A lymphocyte proliferation assay performed on isolated kidney leukocytes recovered from sh at different time points following DNA immunization, showed that antigen-specic restimulation had a kinetic very similar to that of the antibody production [29]. Indirect evidence of cellular immune responses has also been reported. For example, results showing activation of Mx gene expression in muscle after IM administration of antigen-coding but not control plasmid [26], suggest the production of interferon in association with an antigen-specic response. Mx genes encode interferon-inducible proteins that confer nonspecic protection against viral infection in mammals [61].

38

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

They have also been described in sh, although their antiviral function has not yet been established [62]. Furthermore, upregulation of MHC class II expression by DNA vaccines has also been shown in sh, at the site of injection [26]. Expression of these genes is indicative of the recruitment of activated immunocompetent cells, such as B-cells, macrophages and activated T-cells, that are associated with MHC II expression [55]. Such cells could display antigenic peptides and provide help to both B and T lymphocytes. Detection of specic CTL in vaccinated sh has not yet been possible, but there is some indirect evidence indicating that DNA immunization also induces this type of immune response. Fish muscle co-injected with plasmids encoding luciferase (a nonantigenic reporter gene) and VHSV-G protein had greatly decreased luciferase activity by day 28 after DNA injection, compared to control muscles coinjected with luciferase-encoding DNA and a nonexpressing control plasmid [21]. This apparent loss of antigen-expressing muscle cells is immune mediated since it occurs in normal but not immunoincompetent (SCID) mice [46]. In sh, these results suggest the induction of a VHSV-G specic CTL response [21]. Moreover, a DNA vaccine encoding the nucleocapsid N gene of VHSV was also shown to provide protection against live virus in the absence of neutralizing activity in sera from vaccinated sh [22], suggesting that cellular immune responses might be responsible, at least in part, for the protection.

models is evident. In fact, passive immunization of small rainbow trout (700 degree-day old) with serum from adult sh previously immunized with DNA encoding the G gene of IHNV or VHSV provided 100% protection against challenge of the host sh with the respective virus [26].

5. Safety of DNA vaccines for sh Concerns about the safety of DNA vaccines are very different for food source animals and humans [63] and are also somewhat different for sh and other farmed animals. For example, in food source animals there is less concern about integration of DNA into the genome (except in germ cells) and the possibility to induce autoimmune diseases. In mammals, DNA vaccines can induce production of antiDNA antibodies, but these are mostly against singlestranded DNA and are non-pathogenic. Indeed, autoimmune reactions against the hosts DNA have never been observed, even in lupus-prone strains of animals [6466]. Integration of plasmid DNA into the host genome is very unlikely, and the risk of an insertional mutagenic event, which is even less, has been estimated to be 1000 times lower than the rate of spontaneous mutation [67]. Although the potential for integration remains a concern for humans [68], it is less worrisome for sh and other farmed animals with short life spans. Exceptions would be stable integration into germ cells, especially in sh that will be used for reproduction or wild-stock enhancement. Integration of plasmid DNA in the host genome has not been demonstrated in sh yet [25,29], and the presence of the transfected gene product or DNA itself in gonads or eggs from injected sh has not been assessed.

4.3. Protection against live challenge

Although immune responses to DNA vaccines in sh are yet poorly characterized, it is known that they can offer very good protection in small scale challenge trials. Rainbow trout vaccinated with DNA expressing the G protein of infectious hematopoietic necrosis virus (IHNV) had a relative percent survival (RPS) of 75 with live challenge [20]. Similarly, trout DNA-vaccinated against VHSV had RPS values up to 97 and 78, depending on the vaccine dose, after homologous and heterologous challenge respectively [21,22,33]. While it is likely that both humoral and cellmediated immune responses are induced, the importance of antibodies alone for the IHNV and VHSV

5.1. Fate of injected DNA in sh

The persistence of DNA after IM injection in sh differs somewhat from that in mammals. Unlike ndings with mice, where degradation of the plasmid DNA starts only a few hours after injection [69], DNA administered to sh is more stable, with relatively little degradation being detected by Southern blotting. Plasmid could be detected in injected muscle even at the latest time point tested, that is 63

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

39

days after injection [25]. With the polymerase chain reaction (PCR), plasmid DNA was also detected in injected muscle at all time point tested (the latest being 70 days after injection), but not in other tissue extracts [26,29]. The persistence of plasmid DNA is associated with long term expression of the encoded antigen [29].

5.2. Safety for the consumer

The only potential risk to the consumer is that associated with ingestion of plasmid DNA remaining in the injected esh. In mice, it has been estimated that 1 to 2% of orally ingested plasmid DNA survive the gastrointestinal tract and persist as short fragments in the nuclei of intestinal epithelia and Peyers patches cells, in the peripheral white blood cells and in the cells of the spleen and liver [70,71]. In spleen cells, plasmid fragments have been found in rare instances to be covalently linked to DNA with a high degree of homology to mouse genes [71]. Moreover, DNA orally administered to pregnant mice (50 mg of plasmid daily for 12 weeks) was found to cross the placenta and was subsequently detected in various organs of fetal and newborn animals, but the association with abnormalities has not been investigated [72]. It must be realized, however, that the amounts of DNA given to these mice greatly exceed what might be ingested by eating a vaccinated sh. Most (if not all) of the plasmid DNA injected into the sh would have been degraded by the time the esh is eaten, some months or years later. Even if the sh were eaten immediately after injection, the amount of plasmid DNA consumed would be extremely minute compared to the total amount of nucleic acid from all sources, including bacteria and viruses, normally ingested daily. Moreover, licensed veterinary vaccines made of inactivated or attenuated pathogens already contain large amount of nucleic acid, mostly from the genome of the pathogen, as a normal component. From this point of view, DNA vaccines should be no more dangerous than currently used vaccines.

more, DNA that survives in transfected cells will be degraded when these cells are destroyed by antigenspecic CTLs. Nevertheless, a small proportion of DNA can survive for prolonged periods in injected muscle tissue of sh. The prolonged expression of foreign genes was initially thought to be a risk for tolerization to the gene product. This would affect not only the immunized sh, but also other sh, since tolerant animals could become carriers and continuously shed infectious agents into the environment. However, tolerization has never been demonstrated in DNA-immunized animals for any of the many animal models of DNA vaccines developed to date [23]. Moreover, any amount of antigen secreted into an aqueous environment by putative tolerized animals would be extremely small and immediately diluted. The potential risks for the environment, when vaccinated sh are released or escape from farms, are extremely minute. DNA injected sh are not transgenic animals, and it would be very unlikely that they could pass on the plasmid, or part of it, to other organisms or to the next generation of sh. Therefore, the risk of integration in other organisms is extremely small and even if did occur, probabilities for a mutagenic event to happen would be even lower, with the risk of an integrative event in a stem cell being innitesimally small, as discussed previously. Again, risks are unlikely to be greater than with current widely used vaccines.

6. Perspectives for DNA vaccines in the aquaculture industry Results obtained with the rst experiments on DNA vaccines for sh are very encouraging, but several issues still have to be addressed before it will be feasible to have large-scale application of this technology on sh farms. Administration by injection has been shown to be very efcient, and will likely be the rst method developed for commercial DNA immunization of sh. However, alternative routes to deliver nucleic acid vaccines are required for very small sh. Oral or immersion vaccination would be more appropriate in this case. These methods are also desirable for sh of lower value, such as channel catsh, for which the cost of

5.3. Risks for other sh and the environment

Most of the injected plasmids that fail to enter the nucleus of cells are rapidly degraded by nucleases in the extracellular space and in the cytoplasm. Further-

40

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43

carrying out the injection procedure would be prohibitive. Vaccines are normally not administered before the sh are immunocompetent, although DNA vaccines have performed well in neonates of other species [7376]. This is possibly due to the fact that antigen expression can continue until the immune system is mature enough to respond to it [23]. Protection in sh only needs to last long enough for animals to reach market size or for some diseases, a size where the pathogen is no longer a risk. Long-term protection by DNA vaccines still has to be demonstrated in sh, and this will be most important for diseases affecting adult animals. There are also certain requirements to be met prior to the commercial application of DNA vaccines. The most important is the need to meet specic regulations from government agencies. It is also desirable to gain public acceptation of the technology, especially by sh farmers, who will be the actual customers for the vaccine, and by the consumers who will eat the sh. While DNA vaccines are generally inexpensive to manufacture, it may be possible to further reduce the cost of vaccine delivery to sh. For example, plasmids could be added to existing multivalent protein vaccines. It has already been shown that DNA administered IP with an oil adjuvant, similarly to most current sh vaccines, induced nearly as high antibody production as pure plasmid injected IM [28]. Mixing protein and DNA vaccines has been found to be effective in mammals [H.J. Baker and B.F. Smith, personal communication]. Different plasmids may also be coinjected without losing immunogenicity of either components [26], although colinear expression with more than one gene on a single plasmid is more cost effective from a manufacturing point of view. Finally, improvement of the vector backbone to express greater amounts of the foreign protein or to have stronger immune responses against the antigen may also allow lower doses of DNA to be used and thus further reduce the cost.

model antigens give very promising results, with induction of both humoral and cellular immune responses. Moreover, their efcacy in challenge with live pathogens has been demonstrated with two important viral diseases for which no commercial vaccines are available. The increasing need for effective sh vaccines in the rapidly growing aquaculture industry is a driving force for the development of new products. DNA vaccines could circumvent many of the limitations associated with traditional methods of vaccination, but in the end, their usefulness will depend upon a favorable balance of efcacy, safety and cost [77]. The rst two factors should not be a problem for DNA vaccines, but the cost efciency still has to be demonstrated for low-valued species of sh.

References
[1] F.P. Meyer, Aquaculture disease and health management, J. Anim. Sci. 69 (1991) 42014208. [2] D.T. Hanfman, The Status and Potential of Aquaculture in the United States: an Overview and Bibliography, U.S. Department of Agriculture, National Agricultural Library, 1993. [3] P.M. Aasjord, E. Slinde, Fish vaccine: development, production and use of bacterial vaccines, with special reference to salmon, in: A.M. Martin (Ed.), Fisheries Processing: Biotechnological Applications, Chapman & Hall, London, 1994, pp. 432465. [4] E. Mialhe, E. Bachere, V. Boulo, J.P. Cadoret, C. Rousseau, V. CedeZo, E. Saraiva, L. Carrera, J. Calderon, R.R. Colwell, Future of biotechnology-based control of disease in marine invertebrates, Mol. Mar. Biol. Biotechnol. 4 (1995) 275 283. [5] J.C. Leong, J.L. Fryer, Viral vaccines for aquaculture, Annual Rev. Fish Dis. 3 (1993) 225240. [6] C.B. Munn, The use of recombinant DNA technology in the development of sh vaccines, Fish Shellsh Immunol. 4 (1994) 459473. [7] C.McL. Press, A. Lillehaug, Vaccination in European salmonid aquaculture: a review of practices and prospects, Br. Vet. J. 151 (1995) 4569. [8] S.G. Newman, Bacterial vaccines for sh, Ann. Rev. Fish Dis. 3 (1993) 145185. [9] R.A. Schnick, D.J. Alderman, R. Armstrong, R. Le Gouvello, S. Ishihara, E.C. Lacierda, S. Percival, M. Roth, World wide aquaculture drug and vaccine registration progress, Bull. Eur. Ass. Fish Pathol. 17 (1997) 251260. [10] P.J. Midtlyng, Novel vaccines & new vaccination strategies for sh, Bull. Eur. Fish Pathol. 17 (1997) 239244. [11] J.C. Leong, E. Anderson, L.M. Bootland, P.-W. Chiou, M.

7. Conclusions The development of DNA vaccines for sh is still in an early stage. Nevertheless, experiments with

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43 Johnson, C. Kim, D. Mourich, G. Trobridge, Fish vaccine antigens produced or delivered by recombinant DNA technologies, Dev. Biol. Stand. 90 (1997) 267277. J.R. Winton, Molecular approaches to sh vaccines, J. Appl. Ichthyol. 14 (1998) 153158. G.J. Cox, T.J. Zamb, L.A. Babiuk, Bovine herpesvirus 1: immune responses in mice and cattle injected with plasmid DNA, J. Virol. 67 (1993) 56645667. H.L. Davis, M.-L. Michel, R.G. Whalen, DNA-based immunization induces continuous secretion of hepatitis B surface antigen and high levels of circulating antibody, Hum. Mol. Genet. 2 (1993) 18471851. J.B. Ulmer, J.J. Donnelly, S.E. Parker, G.H. Rhodes, P.L. Felgner, V.J. Dwarki, S.H. Gromkowski, R.R. Deck, C.M. DeWitt, A. Friedman, L.A. Hawe, K.R. Leander, D. Martinez, H.C. Perry, J.W. Shiver, D.L. Montgomery, M. Liu, Heterologous protection against inuenza by injection of DNA encoding a viral protein, Science 256 (1993) 1745 1749. B. Wang, K.E. Ugen, V. Srikantin, M.G. Agadjanyan, K. Dang, Y. Fefaeli, A.I. Sato, J. Boyer, W.V. Williams, D.B. Weiner, Gene inoculation generates immune responses against human immunodeciency virus type I, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 41564160. M. Sedegah, R. Hedstrom, P. Hobart, S.L. Hoffman, Protection against malaria by immunization with plasmid DNA encoding circumsporozoite protein, Proc. Natl. Acad. Sci. U.S.A. 91 (1994) 98669870. Z.Q. Xiang, S. Spitalnik, M. Tran, W.H. Wunner, J. Cheng, H.C. Ertl, Vaccination with a plasmid vector carrying the rabies virus glycoprotein gene induces protective immunity against rabies virus, Virology 199 (1994) 132140. E. Hansen, K. Fernandes, G. Goldspink, P. Butterworth, P.K. Umeda, K.-C. Chang, Strong expression of foreign genes following direct injection into sh muscle, FEBS Lett. 290 (1991) 7376. E.D. Anderson, D.V. Mourich, S.C. Fahrenkrug, S. LaPatra, J. Shepherd, J.C. Leong, Genetic immunization of rainbow trout (Oncorhynchus mykiss) against infectious hematopoietic necrosis virus, Mol. Mar. Biol. Biotechnol. 5 (1996) 114 122. J. Heppell, N. Lorenzen, N.K. Armstrong, T. Wu, E. Lorenzen, K. Einer-Jensen, J. Schorr, H.L. Davis, Development of DNA vaccines for sh: vector design, intramuscular injection and antigen expression using viral haemorrhagic septicaemia virus genes as model, Fish Shellsh Immunol. 8 (1998) 271286. N. Lorenzen, E. Lorenzen, K. Einer-Jensen, J. Heppell, T. Wu, H.L. Davis, Protective immunity to VHS in rainbow trout (Oncorhynchus mykiss, Walbaum) following DNA vaccination. Fish Shellsh Immunol. 8 (1998) 261270. H.L. Davis, M.J. McCluskie, DNA vaccines for viral diseases, Microbes and Infections 1 (1999) 721. J.J. Donnelly, J.B. Ulmer, J.W. Shiver, M.A. Liu, DNA vaccines, Ann. Rev. Immunol. 15 (1997) 617648. E.D. Anderson, D.V. Mourich, J.C. Leong, Gene expression in rainbow trout (Oncorhynchus mykiss) following in-

41

[26]

[12] [13]

[14]

[27]

[15]

[28]

[29]

[30]

[16]

[31]

[17]

[32]

[18]

[33]

[19]

[34]

[20]

[35]

[21]

[36]

[37]

[22]

[23] [24] [25]

[38]

[39]

tramuscular injection of DNA, Mol. Mar. Biol. Biotechnol. 5 (1996) 105113. P. Boudinot, M. Blanco, P. de Kinkelin, A. Benmansour, Combined DNA immunization with the glycoprotein gene of viral hemorrhagic septicemia virus and infectious hematopoietic necrosis virus induces double-specic protective immunity and nonspecic response in rainbow trout, Virology 249 (1998) 297306. P.H. Russell, T. Kanellos, I.D. Sylvester, K.C. Chang, C.R. Howard, Nucleic acid immunisation with a reporter gene results in antibody production in goldsh (Carassius auratus L.), Fish Shellsh Immunol. 8 (1988) 121128. T. Kanellos, I.D. Sylvester, C.R. Howard, P.H. Russell, DNA is as effective as protein at inducing antibody in sh, Vaccine 17 (1999) 965972. T. Kanellos, I.D. Sylvester, A.G. Ambali, C.R. Howard, P.H. Russell, The safety and longevity of DNA vaccines for sh, Immunology 96 (1999) 307313. P.M. Schulte, D.A. Powers, M. Schartl, Efcient gene transfer into Xiphophorus muscle and melanoma by injection of supercoiled plasmid DNA, Mol. Mar. Biol. Biotechnol. 7 (1998) 241247. A. Rahman, N. Maclean, Fish transgene expression by direct injection into sh muscle, Mol. Mar. Biol. Biotechnol. 1 (1992) 286289. M. Gomez-Chiarri, S.K. Livingston, C. Muro-Cacho, S. Sanders, R.P. Levine, Introduction of foreign genes into the tissue of live sh by direct injection and particle bombardment, Dis. Aquat. Org. 27 (1996) 512. N. Lorenzen, E. Lorenzen, K. Einer-Jensen, J. Heppell, H.L. Davis, Genetic vaccination of rainbow trout against VHS virus: Small amounts of DNA protect against a heterologous serotype, Virus Res. (1999) in press. B.M. Bearzotti, E. Perrot, C. Michard-Vanhee, G. Jolivet, J. Attal, M.-C. Theron, C. Puissant, M. Dreano, J.J. Kopchick, R. Powell, F. Gannon, L.-M. Houdebine, D. Chourrout, Gene expression following transfection of sh cells, J. Biotechnol. 26 (1992) 315325. T.S. Kanellos, I.D. Sylvester, V.L. Butler, A.G. Ambali, C.D. Partidos, A.S. Hamblin, P.H. Russell, Mammalian granulocytemacrophage colony-stimulating factor and some CpG motifs have an effect on the immunogenicity of DNA and subunit vaccines in sh, Immunology 96 (1999) 507510. J.J. Donnelly, D. Martinez, K.U. Jansen, R.W. Ellis, D.L. Montgomery, M.A. Liu, Protection against papillomavirus with a polynucleotide vaccine, J. Infect. Dis. 173 (1996) 314320. M.-L. Michel, H.L. Davis, M. Schleef, M. Mancini, P. Tiollais, R.G. Whalen, DNA-mediated immunization to the hepatitis B surface antigen in mice: aspects of the humoral response mimic hepatitis B viral infection in humans, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 53075311. D.M. Klinman, G. Yamshchikov, Y. Ishigatsubo, Contribution of CpG motifs to the immunogenicity of DNA vaccines, J. Immunol. 158 (1997) 36353639. A.M. Krieg, A.-K. Yi, S.J. Chorr, H.L. Davis, The role of CpG dinucleotides in DNA vaccines, Trends Microbiol. 6 (1998) 2326.

42

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43 mediated immunization to hepatitis B surface antigen: longevity of primary response and effect of boost, Vaccine 14 (1996) 910915. D.M. Justewicz, M.J. Morin, H.L. Robinson, R.G. Webster, Antibody-forming cell response to virus challenge in mice immunized with DNA encoding the inuenza virus hemagglutinin, J. Virol. 69 (1995) 77127717. R.R. Deck, C.M. DeWitt, J.J. Donnelly, M.A. Liu, J.B. Ulmer, Characterization of humoral immune responses induced by an inuenza hemagglutinin DNA vaccine, Vaccine 15 (1997) 7178. S.L. Kaattari, J.D. Piganelli, The specic immune system: humoral defense, in: G. Iwama, T. Nakanishi (Eds.), The Fish Immune System: Organism, Pathogen, and Environment, Academic Press, San Diego, 1996, pp. 207254. D.J. Lee, M. Corr, D.A. Carson, Control of immune responses by gene immunization, Ann. Med. 30 (1998) 460 468. J. Pavlovic, P. Staeheli, The antiviral potentials of Mx proteins, J. Interferon Res. 11 (1991) 215219. J.C. Leong, G.D. Trobridge, C.H.Y. Kim, M. Johnson, B. Simon, Interferon-inducible Mx proteins in sh, Immunol. Rev. 166 (1998) 349363. K. Yamanouchi, T. Barrett, C. Kai, New approaches to the development of virus vaccines for veterinary use, Rev. Sci. Tech. Off. Int. Epiz. 17 (1998) 641653. D.S. Pisetsky, Immunologic consequences of nucleic acid therapy, Antisens Res. Develop. 5 (1995) 219225. G. Mor, M. Singla, A.D. Steinberg, S.L. Hoffman, K. Okuda, D.M. Klinman, Do DNA vaccines induce autoimmune Disease?, Hum. Gene Ther. 8 (1997) 293300. D.S. Pisetsky, DNA and the immune system, Ann. Internal Med. 126 (1997) 169171. W.W. Nichols, B.J. Ledwith, S.V. Manam, P.J. Troilo, Potential DNA vaccine integration into host cell genome, Ann. N.Y. Acad. Sci. 772 (1995) 3039. F.R. Vogel, N. Sarver, Nucleic acid vaccines, Clin. Microbiol. Rev. 8 (1995) 406410. J.A. Wolff, P. Williams, G. Ascadi, S. Jiao, A. Jani, C. Chong, Conditions affecting direct gene transfer into rodent muscle in vivo, BioTechniques 4 (1991) 474485. R. Schubbert, C. Lettmann, W. Doerer, Ingested foreign (phage M13) DNA survives transiently in the gastrointestinal tract and enters the bloodstream of mice, Mol. Gen. Genet. 242 (1994) 495504. R. Schubbert, D. Renz, B. Schmitz, W. Doerer, Foreign (M13) DNA ingested by mice reaches peripheral leukocytes, spleen and liver via the intestinal wall mucosa and can be covalently linked to mouse DNA, Proc. Natl. Acad. Sci. U.S.A. 94 (1997) 961966. R. Schubbert, U. Hohlweg, D. Renz, W. Doerer, On the fate of orally ingested foreign DNA in mice: chromosomal association and placental transmission to the fetus, Mol. Gen. Genet. 259 (1998) 569576. A. Bot, S. Bot, A. Garcia-Sastre, C. Bona, DNA immunization of newborn mice with a plasmid-expressing nucleoprotein of inuenza virus, Viral Immunol. 9 (1996) 207210.

[40] A.M. Krieg, T. Wu, R. Weeratna, S.M. Eer, L. LoveHoman, L. Yang, A.-E. Yi, D. Short, H.L. Davis, Sequence motifs in adenoviral DNA block immune activation by stimulatory CpG motifs, Proc. Natl. Acad. Sci. U.S.A. 95 (1998) 1263112636. [41] A.M. Krieg, CpG DNA: a novel immunomodulator, Trends Micobiol. 7 (1999) 6465. [42] J. Heppell, J. Sanchez-Dardon, A.M. Krieg, H.L. Davis, Stimulation of rainbow trout immune cells with oligodeoxynucleotides bearing CpG motifs, in: Proc. 3rd Int. Symp. Aquatic Ani. Health, 1998, p. 101. [43] A.E. Ellis, Recent development in oral vaccine delivery systems, Fish Pathol. 30 (1995) 293300. [44] A.E. Ellis, in: A.E. Ellis (Ed.), Fish Vaccination, Academic Press, New York, 1988, p. 255. [45] A.D. Cohen, J.D. Boyer, D.B. Weiner, Modulating the immune response to genetic immunization, Faseb J. 12 (1998) 16111626. [46] H.L. Davis, C.L. Brazolot Millan, S.C. Watkins, Immunemediated destruction of transfected muscle bers after direct gene transfer with antigen-expressing plasmid DNA, Gene Ther. 4 (1997) 181188. [47] J.A. Wolff, J.J. Ludtke, G. Acsadi, P. Williams, A. Jani, Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle, Hum. Mol. Genet. 1 (1992) 363369. [48] T.T. Poppe, O. Breck, Pathology of Atlantic salmon Salmo salar intraperitoneally immunized with oil-adjuvanted vaccine, A case report. Dis. Aquat. Org. 29 (1997) 219226. [49] A.G. Zapata, A. Chiba, A. Varas, Cells and tissues of the immune system of sh, in: G. Iwama, T. Nakanishi (Eds.), The Fish Immune System: Organism, Pathogen, and Environment, Academic Press, San Diego, 1996, pp. 162. [50] M. Ototake, G.K. Iwama, T. Nakanishi, The uptake of bovine serum albumin by the skin of bath-immunised rainbow trout Oncorhynchus mykiss, Fish Shellsh Immunol. 6 (1996) 321333. [51] J.D. Moore, M. Ototake, T. Nakanishi, Particulate antigen uptake during immersion immunisation of sh: the effectiveness of prolonged exposure and the roles of skin and gill, Fish Shellsh Immunol. 8 (1998) 393407. [52] R.A. Robohm, R.A. Koch, Evidence for oral ingestion as the principal route of antigen entry in bath-immunized sh, Fish Shellsh Immunol. 5 (1995) 137150. [53] C.J. Secombes, The nonspecic immune system: cellular defenses, in: G. Iwama, T. Nakanishi (Eds.), The Fish Immune System: Organism, Pathogen, and Environment, Academic Press, San Diego, 1996, pp. 63104. [54] T. Yano, The nonspecic immune system: humoral defense, in: G. Iwama, T. Nakanishi (Eds.), The Fish Immune System: Organism, Pathogen, and Environment, Academic Press, San Diego, 1996, pp. 105158. [55] M.J. Manning, T. Nakanishi, The specic immune system: cellular defenses, in: G. Iwama, T. Nakanishi (Eds.), The Fish Immune System: Organism, Pathogen, and Environment, Academic Press, San Diego, 1996, pp. 159205. [56] H.L. Davis, M. Mancini, M.-L. Michel, R.G. Whalens, DNA-

[57]

[58]

[59]

[60]

[61] [62]

[63]

[64] [65]

[66] [67]

[68] [69]

[70]

[71]

[72]

[73]

J. Heppell, H.L. Davis / Advanced Drug Delivery Reviews 43 (2000) 29 43 [74] A. Bot, S. Antohi, S. Bot, A. Garcia-Sastre, C. Bona, Induction of humoral and cellular immunity against inuenza virus by immunization of newborn mice with a plasmid bearing a hemagglutinin gene, Intl. Immunol. 9 (1997) 16411650. [75] D.E. Hassett, J. Zhang, J.L. Whitton, Neonatal DNA immunization with a plasmid encoding an internal viral protein is effective in the presence of maternal antibodies and protects

43

against subsequent viral challenge, J. Virol. 71 (1997) 7881 7888. [76] A.M. Prince, R. Whalen, B. Brotman, Successful nucleic acid based immunization of newborn chimpanzees against hepatitis B virus, Vaccine 15 (1997) 916919. [77] C.W. Beard, P.W. Mason, Out on the farm with DNA vaccines, Nature Biotechnol. 16 (1998) 13251328.