Construction of Recombinant Vaccinia Virus Vectors

The development of recombinant vaccinia viruses depended on a method of introducing a foreign gene into the vaccinia virus genome. Homologous recombination between poxviruses was well known and had been demonstrated by coinfecting cells with two viruses ¹⁶⁶ and by infecting cells with one virus and transfecting them with genomic DNA ^{167, 168} or cloned DNA segments. ¹⁶⁹ It is likely that DNA recombination occurs in the cytoplasm by enzymes encoded by vaccinia virus. Less well understood at the time was how to achieve expression of foreign genes. The recognition of vaccinia virus promoter elements provided a general method of preparing vaccinia virus expression vectors ^{161, 170, 171} that is illustrated in Figure 6-9. Insights achieved through basic studies of vaccinia virus promoters have led to substantial improvements in the level of gene expression. ^{172, 173} Other innovations, including alternative methods of selecting or identifying recombinant vaccinia viruses and the insertion of foreign genes by direct ligation, are summarized elsewhere. ¹⁷⁴

Figure 6-9

The insertion plasmid (or transfer vector) contains restriction endonuclease sites for ligation of the complete open reading frame (ATG----------TAA) of a foreign gene adjacent to a vaccinia virus promoter as well as flanking vaccinia virus DNA sequences (more...)

Selection of a Vaccinia Virus Strain

The WR strain of vaccinia virus, favored for basic poxvirus research in the United States and widely used to make recombinant viruses for laboratory studies, is unsuitable for vaccines. The four vaccinia virus strains administered most often for smallpox vaccination were EM-63, Lister, New York City Board of Health, and Temple of Heaven. Of these, the New York City Board of Health strain had relatively low pathogenicity ¹⁷⁵ and was chosen to make a recombinant vaccinia virus intended for human use. Although the latter appeared to be safe in a small clinical trial, ¹⁷⁶ further attenuation of recombinant vaccinia viruses seems prudent for large-scale administration. Several approaches have been taken to achieve a safer vector.

Although 50 or more of the nearly 200 genes of vaccinia virus are dispensable for replication in tissue culture cells, the deletion of some of these genes reduces virulence in animal models. ^{177– 179} The deletional approach to making a safe vector was exemplified by the removal of 18 genes from the Copenhagen strain of vaccinia virus, thereby producing a highly attenuated derivative called NYVAC. ¹⁸⁰ Several studies have indicated that NYVAC has good potential for human and veterinary vaccines. ^{181, 182} An alternative approach was to use one of several highly attenuated strains of vaccinia virus that were derived by serial passages in vitro and had been tested in humans before the eradication of smallpox. ¹⁷⁵ One of these strains, MVA, suffered multiple deletions and became severely host range restricted during more than 500 passages in chick embryo cells, providing a high degree of attenuation. ^{183, 184} Laboratory studies demonstrated unimpaired MVA gene expression in human cells and a block in virion morphogenesis. ¹⁸⁵ The ability to achieve high expression of recombinant genes despite abortive replication is a remarkable feature of this mutant virus. Even though MVA probably does not replicate significantly in animal models, excellent immune responses to recombinant proteins were obtained; moreover, the dose required was similar to that of a standard replication-competent strain. ^{186– 188}

Vaccinia Virus as a Tool for Vaccine Research

Recombinant vaccinia viruses provide a powerful means of dissecting the immune responses of humans and experimental animals to individual gene products of infectious agents. Only a few examples can be mentioned here. Thus, recombinant vaccinia viruses were used to demonstrate that the HA and NP proteins of influenza virus induced subtype-specific and cross-reactive cytotoxic T-cell responses, respectively. ^{189, 190} Evidence of human immunodeficiency virus type 1 (HIV-l)-specific cytotoxic T cells in patients with acquired immunodeficiency syndrome (AIDS) was first obtained by use of recombinant vaccinia viruses expressing the envelope or internal proteins to prepare target cells. ^{191, 192} Indeed, recombinant vaccinia viruses have become an important tool for cellular immunologists. ¹⁹³ Because proteins expressed in mammalian cells by recombinant vaccinia viruses are folded, processed, and transported normally, they can be used to either induce or bind antibodies that recognize conformational epitopes. ¹⁹⁴

The wide host range of vaccinia virus makes it possible to determine protective immune responses against infectious agents in a variety of experimental animals from rodents to primates. For example, the F glycoprotein is most important for inducing protection to respiratory syncytial virus, ¹⁹⁵ whereas the HN protein is better for parainfluenza virus type 3 ¹⁹⁶ and type 5. ¹⁹⁷ Protection elicited by the syncytial virus M2 protein is due to CD8⁺ T cells, whereas that induced by the F and G proteins is due to antibodies. ¹⁹⁸ Similar results, with respect to the HA and NP proteins, have been obtained in studies with influenza virus. ¹⁹⁹ In some cases, vaccination has a priming effect that is followed by an anamnestic antibody response, as indicated for the protection of chimpanzees after inoculation with a recombinant vaccinia virus expressing the hepatitis B surface antigen. ²⁰⁰ A list of viruses for which protective immune responses have been obtained may be found in a review. ²⁰¹

The induction of strong cytotoxic T-cell responses elicited by recombinant vaccinia viruses has led to their evaluation as tumor vaccines in model systems. ^{202– 205}

Other Poxvirus Vectors

The procedures developed for the construction of recombinant vaccinia viruses have been applied to members of other poxvirus genera including avian poxviruses ²⁰⁶ and capripoxviruses. ²⁰⁷ Although the avian poxviruses are naturally host range restricted, gene expression and protective immunity can be established in nonavian species. ^{208, 209} As nonreplicating vectors, avian poxviruses should be exceptionally safe recombinant vaccines.

Human Vaccines

Although vaccinia virus vectors have proved extremely useful for vaccine research as well as for research in many other fields, the potential for human vaccines is still under investigation. As for all vaccines, the critical factors include safety and efficacy as well as the facility for vaccine production, distribution, and administration. In addition, there are special questions regarding prior immunity to vaccinia virus acquired either through smallpox vaccination or through a recombinant vaccine and the design of polyvalent vaccines.

Safety issues have been minimized by the demonstration that host-restricted or “nonreplicating” vaccinia virus vectors, such as the MVA and NYVAC, are immunogenic. At the National Institutes of Health intramural laboratories in Bethesda, research with both of these strains is permitted at biosafety level 1 conditions, whereas level 2 and a recent smallpox vaccination are required for working with standard vaccinia virus strains.

Generic and specific factors are involved in vaccine efficacy. With regard to the former, great improvements in gene expression have been made so that present generation vectors produce many times more recombinant protein than the original vectors. In some instances, immunogenicity has been improved by altering the presentation of the recombinant protein so that it is plasma membrane associated rather than intracellular or secreted. ^{210, 211} Promising results have been obtained by constructing recombinant vaccinia viruses that coexpress an immunogen and an immunomodulatory cytokine. ^{212– 214}

Although the use of live attenuated viruses as human vaccines may require no special knowledge regarding the targets of immunity, such specific information is needed for recombinant vectors. For many viruses, the membrane glycoproteins or capsids are targets of neutralizing antibody and the internal proteins provide good targets for cytotoxic T cells. Animal models may be helpful in identifying those targets that provide protective immunity. In addition, some infectious agents have special vaccine requirements such as those related to portal of entry, site of replication, antigenic variation, type of immune response needed, and presence of maternal antibodies, which may or may not be met by a recombinant vaccine.

The smallpox vaccine was most frequently prepared from vaccinia virus that was propagated in the skin of an animal, but an approved, stable, freeze-dried vaccine was produced in monolayer cultures of primary rabbit kidney cells. ²¹⁵ Acceptable cultured cell lines or primary chick embryo fibroblasts would be alternative substrates for virus propagation. Thus, there should be no impediment to the preparation of vaccines that meet present standards of purity. Presumably, procedures for freeze-drying could be adapted to the production of recombinant vaccinia viruses, and it is hoped that such preparations would retain the excellent thermal stability that made a cold chain unnecessary for the smallpox vaccine.

The smallpox vaccine was generally administered by scarification of the skin or less commonly by a high-pressure jet injector. ¹⁷⁵ The intradermal route was used for clinical testing of a recombinant virus made with the New York City Board of Health strain. ¹⁷⁶ However, other routes (e.g., nasal, oral, subcutaneous, or intramuscular) may be preferred for nonreplicating strains of vaccinia virus.

Although it was based on a small sampling, prior smallpox vaccination appeared to diminish the immune response to a recombinant vaccine. ¹⁷⁶ For children and the majority of individuals born during the past 20 years, who have not received a smallpox vaccination, this would not be a problem. However, a poxvirus may not be useful as a carrier for revaccination with a second gene because of the immune response to the vector. Whether prior immunity could be overcome by using vectors that express more recombinant protein or through alternative routes of administration is uncertain. Nevertheless, immunization with a single vector expressing multiple genes, simultaneously with a cocktail of vectors, or successively with distantly related poxvirus vectors might obviate such problems.

Accelerated efforts to develop an AIDS vaccine have led to the human testing of a first-generation recombinant vaccinia virus expressing the HIV-1 envelope gene. ^{176, 216} The modest immune response detected may be due in part to the relatively weak promoter used and the failure to eliminate poxvirus early transcriptional stop signals within the HIV-1 gene. ²¹⁷ The HIV-1 neutralizing antibody response, however, was augmented by secondary immunization with a subunit HIV-1 envelope protein. ^{218, 219} There is considerable enthusiasm for such a prime-boost strategy because it can stimulate cell-mediated and humoral immunity. Prime-boost vaccinations carried out with a recombinant canarypox virus and recombinant HIV-1 envelope protein induced HIV-specific cytotoxic cells and neutralizing antibody in phase I clinical trials. Expanded phase II trials are in progress. ^{220– 222}

A recombinant vaccinia virus that expresses the major Epstein-Barr virus membrane glycoprotein was immunogenic when it was administered to infants and young children and may have delayed or prevented natural infection for a period of 16 months. ²²³ A recombinant canarypox virus expressing the rabies virus glycoprotein was safe in humans, induced functional antibody to rabies glycoprotein, elicited cellular responses to rabies virus, and could be used successfully for boosting at a 6-month interval. ^{224, 225}

Veterinary and Wildlife Vaccines

Substantially different factors are involved in the applicability of vaccines for medical and veterinary practices. ²²⁶ Economic criteria, although of considerable importance for human vaccines in-developing countries, are decisive for most veterinary vaccines. Also, there is far more latitude in the manufacture and use of veterinary vaccines than has been permitted by regulatory agencies for human vaccines. In addition, durable immunity is not important for livestock, and a small number of vaccine-associated illnesses can be tolerated in veterinary vaccines. Because live fowlpox virus vaccines are already used in the poultry industry to prevent fowlpox, recombinant poxvirus vaccines should be practical.

Recombinant vaccinia viruses have been shown to protect animals against diseases of veterinary importance including vesicular stomatitis virus ²²⁷ and rinderpest ²²⁸ in cattle, pseudorables virus in swine, ²²⁹ and influenza virus in chickens. ²³⁰ A recombinant vaccinia virus expressing the rabies virus glycoprotein ²³¹ has been successfully administered in bait form as a wildlife vaccine in both the United States and Europe. ^{232, 233}

Other poxviruses are also being tested as vectors for veterinary applications. Examples include a raccoon poxvirus vector for raccoons against rabies virus ²³⁴ ; a capripoxvirus vector for cattle against rinderpest virus ²³⁵ ; a swinepox virus vector for pigs against pseudorabies virus ²³⁶ ; fowlpox vectors for chickens against influenza virus, ²³⁷ Newcastle disease virus, ^{238, 239} and infectious bursal disease virus ²⁴⁰ ; and canarypox virus for cats against feline leukemia virus. ²⁴¹