The Development and Use of a Poxvirus Vector System to Protect Pigs Against Swine Influenza Virus Patricia Louise White Foley Iowa State University

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

1998 The development and use of a poxvirus vector system to protect pigs against swine influenza virus Patricia Louise White Foley Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Agriculture Commons, Animal Sciences Commons, Immunology and Infectious Disease Commons, Medical Immunology Commons, Microbiology Commons, and the Veterinary Pathology and Pathobiology Commons

Recommended Citation Foley, Patricia Louise White, "The development and use of a poxvirus vector system to protect pigs against swine influenza virus " (1998). Retrospective Theses and Dissertations. 12192. https://lib.dr.iastate.edu/rtd/12192

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

This manuscript has been reproduced from the microfilm master. UMI films the text directly firom the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overiaps. Each original is also photographed in one exposure and Is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning 300 North Zeeb Road, Ann Artx>r, Ml 48106-1346 USA 800-521-0600

The development and use of a poxvirus vector system

to protect pigs against swine influenza virus

Patricia Louise White Foley

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major; Immunobiology

Major Professor: Prem S. Paul

Iowa State University

Ames, Iowa

1998

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103 11

Graduate College Iowa State University

This is to certify that the Doctoral dissertation of

Patricia Louise White Foley has met the dissertation requirements of Iowa State University

Signature was redacted for privacy. Major Professor

Signature was redacted for privacy. ajor Program

Signature was redacted for privacy. For the Graduate College m

TABLE OF CONTENTS

Page LIST OF SELECTED ABBREVIATIONS V ABSTRACT viii

GENERAL INTRODUCTION 1 Dissertation Organization 5 Literature Review 5 Statement of the problem 34 References 3 5

L EVALUATION OF EVIMUNOGENICITY, SAFETY, AND HOST RESTRICTION OF THE HIGHLY ATTENUATED MVA STRAIN OF VACCINIA VIRUS EXPRESSING HUMAN INFLUENZA VIRUS HEMAGGLUTININ AND NUCLEOPROTEIN GENES 48 Abstract 48 Introduction 49 Materials and Methods 52 Results 56 Discussion 59 References 61

2. RECOMBINANT VACCINIA VIRUS CONTAINING HEMAGGLUTININ AND NUCLEOPROTEIN GENES FROM A PORCINE STRAIN OF INFLUENZA VIRUS PROTECTS PIGS AGAINST SWINE INFLUENZA VIRUS 77 Structured Abstract 77 Introduction 79 Materials and Methods 83 Results 89 Discussion 92 References 96

3. EVALUATION OF INHIBITION OF IMMUNE RESPONSE TO VACCINATION FOLLOWING SEQU'ENTIAL USE IN PIGS OF TWO MODIFIED VACCINLA VIRUS ANKARA (MVA) RECOMBINANTS EXPRESSING HETEROLOGOUS INFLUENZA VIRUS GENES 113 Abstract 113 Introduction 114 Materials and Methods 118 Results 124 Discussion 126 References 128 GENERAL CONCLUSIONS References V

LIST OF SELECTED ABBREVIATIONS

AGIO agar gel immunodifilision assay.

CDCD caesarean-derived colostrum-deprived.

CEF chicken embryo fibroblast cells. cELISA competitive enzyme-linked immunosorbent assay.

CPE cytopathic effect.

EID50 mean egg infectious dose.

FFU fluorescent focal units.

GMT geometric mean titer.

HA hemagglutinin, the major surface protein of influenza virus and the target of

influenza virus-neutralizing antibody.

HAI hemagglutination inhibition assay, which measures serum antibody to HA.

HlNl a designation for a type A influenza virus with a hemagglutinin gene of HI

subtype (out of 15) and a neuraminidase gene ofNl subtype (out of 9).

H3N2 another subtype of tj^je A influenza virus.

H1N2 believed to be the result of reassortment of genetic segments of HlNl and

H3N2 influenza viruses.

IA/31 A/Swine/Iowa/31, an HlNl influenza virus of the 'classic' SIV variety,

IM intramuscular (route of inoculation).

EST intranasal (route of inoculation). rN88 A/Swine/In/1726/88, a reference strain of'classic' HlNl SIV isolated at the VI

University of Wisconsin.

IN/88 A/Swine/In/1726/88, a reference strain of'classic' HlNl SIV isolated at the

University of Wisconsin.

ISU Srv an Iowa field isolate of SIV, Iowa State University Veterinary Diagnostic

Laboratory swine isolate #40776.

MN/88 an HlNl influenza virus isolated from turkeys in Minnesota.

MO/87 an HlNl influenza virus isolated from turkeys in Missouri.

MVA modified vaccinia Ankara, highly attenuated strain of vaccinia virus.

MVA-HA-NP recombinant MVA virus expressing HA and NP genes of A/PR/8/34, a human

HlNl strain of influenza virus.

NIVA/PR8 recombinant MVA virus expressing HA and NP genes of A/PR/8/34, a human

HlNl strain of influenza virus.

MVA/SrV recombinant MVA virus expressing HA and NP genes of an Iowa field strain

of HlNl swine influenza virus.

NP nucleoprotein, the major target antigen for cytotoxic T lymphocytes, which

improve influenza virus clearance.

PFU plaque-forming units.

PR8 A/PR/8/34, a human HlNl strain of influenza virus.

PR/8 A/PR/8/34, a human HlNl strain of influenza virus.

SD/86 an HlNl influenza virus isolated from turkeys in South Dakota.

SI swine influenza, a disease generally of high morbidity and low mortality in

pigs. vii

SrV swine influenza virus.

WR nonattenuated Western Reserve strain of vaccinia virus, has the fiill vaccinia

virus host range.

WR-HA-NP a WR recombinant expressing the HA and NP genes of the human HlNl PR8

strain of influenza virus. viii

ABSTRACT

The hypothesis tested in this dissertation is that a poxvirus vector system containing influenza virus immunogens protects pigs against swine influenza virus (SrV)-associated disease. The host range-restricted, highly attenuated, and safety-tested modified vaccinia virus Ankara

(MVA) strain was used as a vector for influenza virus hemagglutinin (HA) and nucleoprotein

(NP) genes from the HlNl A/PR/8/34 (PRS) human isolate. This recombinant virus has previously been shown to protect mice against lethal challenge with PRE. First, the ability of the MVA/PR8 recombinant to protect pigs against SIV was examined. Second, a new recombinant, designated NIVA/SrV, was constructed containing the HA and NP genes from a field isolate of SIV submitted to the Iowa State University Veterinary Diagnostic Laboratory.

The MVA/SIV recombinant was evaluated for immunogenicity in vaccinated pigs subsequently challenged with SFV. Third, the possibility of inhibition of immune response in secondary vaccination, following sequential use of two IvIVA recombinants, was explored using the MVA/PR8 and MVA/SIV constructs. The first study demonstrated that protection afforded by MVA/PR8 against SFV was less than complete, that infection, clinical signs of illness, and viral shedding still occurred, albeit to a lesser degree than in nonvaccinated controls. The second study indicated the marked improvement in protection against SIV when pigs were vaccinated with the MVA/SIV construct. The third study provided evidence that sequential use of such MVA recombinants containing inserts from two strains of influenza virus, given two months apart, still generated appropriate immune responses to the different inserts. 1

GENERAL INTRODUCTION

Swine influenza is a major respiratory problem in swine. Infected herds are common and widespread. Moreover, there are several antigenically distinct subtypes of the causative virus that can produce disease in susceptible pigs. There is only one vaccine available, an inactivated product derived from a single strain, which cannot provide protection against all virulent strains. Immunity against influenza virus largely consists of a strong subtype-specific humoral response to a major surface antigen and a heterotypic cell-mediated response to a major internal protein. A live virus vaccine could generate these elements of protection.

However, the ability of influenza virus to infect numerous species, including humans, and to mutate and to undergo genomic reassortment, dampens enthusiasm for introducing live strains into a susceptible population. This caution regarding inadequate attenuation suggests that use of an innocuous vector to express immunogens of influenza virus in a safe, yet effective manner would be beneficial. Numerous candidate vectors have been considered by researchers in the past, but one family of viruses in particular has been used extensively, the poxviruses. These large, relatively self-sufficient viruses generally induce strong enduring immune response in infected individuals. Also, there are members of this family that are not pathogenic and thus provide safe alternatives as recombinant vaccine vector candidates.

The hypothesis to be tested in this dissertation is that a poxvirus vector system containing influenza virus immunogens protects pigs against swine influenza virus (SIV)- associated disease. Three major objectives have been delineated, and each of these has several subobjectives. 2

The first objective is to determine whether the HA and NP genes fi^om the human influenza virus strain A/PR/8/34 (PR8), when expressed by the modified vaccinia Ankara

(MVA) strain of vaccinia virus, provide adequate protection in pigs against a heterologous

SrV challenge. Subobjectives of this objective are to, first, compare the intramuscular (IM) route of inoculation with intranasal (IN) administration in protecting vaccinates; second, determine whether the MVA strain alone, without insertion of foreign genes, has an effect on the immune response of pigs, as compared to pigs receiving no vaccination prior to challenge; third, determine whether the MVA strain is replication incompetent in certain cells or cell lines of swine, as it has been shown to be in those of other mammals; and, fourth, determine whether the MVA recombinant can pass the mouse safety test required for licensure of live virus vaccines, as described in the United States' Code of Federal Regulations, Chapter 9, part 113.33(a).

The study design for achieving the goals of the first objective and its subobjectives will entail the use of colostrum-deprived caesarean-derived pigs, placed in parental MVA and recombinant treatment groups, and vaccinated accordingly at 21 and 35 days of age. At 49 days of age, all pigs, including nonvaccinated controls, will be oronasally challenged with pathogenic A/Sw/IN/1726/88 (IN88) and monitored daily for 7 days, at which time the pigs will be euthanized. In addition, two porcine cell lines, PK-15 (porcine kidney) and ST (swine testicle), as well as primary swine kidney cells, will be evaluated for their ability to sustain

MVA replication and protein expression. Lastly, nine groups of eight mice each will each receive one of three viruses (MVA parent, MVA recombinant, or the wild type Western

Reserve (WR) strain of vaccinia containing HA and NP inserts) by one of three routes, i.e. J•-> intraperitoneally, subcutaneously, or intracranially. A tenth group will serve as negative controls. The mice will be monitored for clinical signs for 7 days, then undergo postmortem examination.

The second objective is to determine whether an MVA recombinant containing HA and NP genes from a porcine strain of influenza virus, administered either by IM or IN route, is able to protect pigs against a homologous SIV challenge. Subobjectives of the second objective are to, first, use the MVA parental strain to develop a recombinant containing HA and NP from Iowa State University Veterinary Diagnostic Laboratory swine isolate #40776

(ISU SrV); second, evaluate the recombinant's expression of the SIV proteins through immunoplaque assay; and, third, determine whether a cell line such as the Madin-Darby

Canine BCidney (MDCK) line can support SIV growth, reducing the present reliance on embryonated egg culture assays.

To achieve the second group of subobjectives, several steps will be enacted. RNA from the ISU SIV strain will be extracted from allantoic fluid of infected egg culture. The

HA and NP genes will undergo reverse transcription (RT) and poljmierase chain reaction

(PCR) amplification using published sequences to select primers. The products will be cloned, sequenced, then subcloned into the plasmid vector used to develop the first recombinant. The plasmid will then be used to transfect the MVA strain. Briefly, MVA- infected chick embryo fibroblasts will receive plasmid DNA in the presence of a lipophilic reagent, then be passaged several times to amplify recombinants. Subsequent plaques reactive to SIV Ab in colorimetric immunoassays will be isolated and plaque-purified. The selected recombinant (MVA/SIV) will be expanded, then evaluated by immunoassay for HA 4 and NP expression.

The third objective is to determine whether there is immune response inhibition of secondary vaccination following sequential use of the two MVA recombinants, MVA/PR8 and MVA/SrV. Subobjectives of this third objective are to compare the effects of IN versus

IM routes of administration on sequential vaccination, and to determine how rapidly titers decline following administration of the first recombinant. In addition, sequence analysis of the HA and NP genes fi^om the two strains will be performed to determine degree of homology.

The experimental design for achieving the third objective should determine whether antibody titer to the MVA vector generated during initial immunization with the PR8 recombinant can interfere with subsequent immunization using the SFV recombinant,

MVA/SIV. Six pigs will receive the PR8 recombinant IM and six IN at 21 and 35 days of age. Four sentinel pigs will remain unvaccinated. At approximately day 74 and day 88 following the first vaccination, each of the 12 vaccinated pigs, as well as the 4 nonvaccinated pigs, will receive the SIV recombinant, by the IM route. The expectation is that even closely spaced, sequential use of the MVA vector will not result in interference to immune response.

Should the MVA recombinants prove efiBcacious and safe, it would be most convenient to be able to administer them and similar constructs repeatedly.

To accomplish the third set of subobjectives, pig antisera will be evaluated by HAI against ISU SFV and PR8 influenza antigen and by SN (constant virus-varying serum method) against SIV and MVA. Blood samples will be collected every 2 weeks until day 140 of the study, or 8 weeks after the last vaccination given. 5

Dissertation Organization

The dissertation is divided into three papers, corresponding to the three major objectives to be achieved, each with individual subobjectives, and written for submission to particular journals. Each paper is presented as a separate research project, with its own

Introduction and sections on Materials and Methods, Results, and Discussion. However, there will be one Literature Review for all three papers, plus an accompanying reference section, prior to discussion of the individual studies, and one section each on General

Conclusions and References, after the three studies have been presented. The Table of

Contents may be consulted for specific page notation.

Literature Review

Jennerian vaccines In 1798, an English physician by the name of Edward

Jenner published his observations on the immunity against smallpox provided by inoculation with cowpox virus (Jenner 1798). Prior practice had been to inoculate material from mild cases of smallpox, but the consequences of this procedure were not always protective. In too many instances this caused the disease one hoped to prevent. Others were aware that exposure to cowpox seemed to protect against smallpox, well before Tenner's time. But,

Tenner's methodical approach and careful experimentation helped calm public fears and justify community vaccination programs, using a related but nonpathogenic virus to build immunity against a dreaded disease (Cartwright & Biddiss 1972). 6

Since that time, there has been steady progress in searching for and developing such

vaccines. More recently, where candidate vaccines couldn't be found in nature, vaccines have

been produced through the use of innovative biotechnology. For example, two inactivated

and six modified live pseudorabies virus (PRV) vaccines are currently licensed as gene-

deleted vaccines (USDA 1998). Gene-deleted vaccines have an advantage in that deleted gene products can be used as serologic markers to distinguish vaccinated from infected animals (van Oirschot, et al. 1996). Gene-deleted vaccines may also be of low virulence if the deleted gene codes for virulence, e. g., Tk-deleted vaccines of PRV have reduced virulence (Kit, et al. 1985; Kit, et al. 1987). Such vaccines have played a very important role in the national pseudorabies eradication effort.

As we approach the twenty-first century, biotechnology is changing the character of biologies even further. New procedures make it possible to isolate from the genome of a highly virulent virus that portion which can provide protective immunity. Insertion of such genetic material into another 'carrier" virus then allows expression of specific protective genes safely in a vaccine. There are many such live vectored vaccines currently under development and several that have been licensed for use in the United States (U. S.) and Europe. The vectors used for the vaccines licensed in the U. S. include fowlpox virus, canarypox virus, and vaccinia virus (USDA 1998). Fowlpox virus recombinant vaccines carry genes from avian influenza virus (Webster, et al. 1991; Webster, et al. 1996) and Newcastle disease virus

(Taylor, et al. 1996) for use in poultry. Canarypox virus has been used as a vector for canine distemper virus genes for disease protection in dogs (Taylor, et al. 1992). Vaccinia virus expressing the rabies glycoprotein (G) gene has been used as an oral vaccine for prevention 7 and control of rabies in wildlife. It has been used successfully to protect foxes in Europe and raccoons in the United States against epizootic rabies (Brochier, et aL 1990; Desmettre, et al.

1990). The vaccine appears to be heat-stable enough to survive distribution by airdrop from helicopter. The vaccine is also safe, having been tested in over 30 species, as well as eflBcacious, to make the possibility of eradication of sylvatic rabies quite real.

Advantages of recombinant vaccines Many highly pathogenic viruses, for which vaccines would be desirable, are too proficient at producing disease to be administered in a live product. For example, few if any whole-rabies virus vaccines are sufBciently attenuated to warrant administering them live into a susceptible species. Although there are many licensed rabies vaccines, they are all inactivated products, with one exception of the recombinant vaccinia virus vaccine for wildlife. In addition, even when a virus strain does not itself produce disease, it may have characteristics that, when recombined with another strain encountered in the field, result in a far worse disease outbreak. Such is the case with influenza viruses, which have a segmented genome capable of reassortment. If an attenuated vaccine strain, with certain gene segments that give it the ability to replicate well in a given host species, should exchange segments with another virulent strain that does not replicate well in that species, the outcome may be a new strain that is both virulent and well-adapted to the new host. That was the fear that characterized the 1997 Hong Kong outbreak in humans of an avian strain of influenza (Claas, et al. 1998b; Suarez, et aL 1998). Fortunately, although the avian strain was able to colonize certain humans, it did not replicate well enough in that host to cause an epidemic in what would have been a completely susceptible. 8 immunologically naive population.

Among the veterinary influenza products, there are eight avian vaccines manufactured by one company, one swine vaccine, and five types of equine vaccines produced by six different companies (USDA 1998). They are all killed virus products, as are the human influenza vaccines used in annual vaccination programs. Human vaccines are changed annually to reflect the newly emerging strains and frequently contain 3 distinct hemagglutinin antigens from diverse strains to maximize protection against disease (Couch, et al. 1996).

One problem with inactivated influenza vaccines is that the immunity generated is only partial (McMichael, et al. 1983). In the presence of a strong adjuvant, antigens can stimulate B-cells and induce a good humoral response. However, there is little cell-mediated immunity generated by killed product, and this can mean the difference between disease and protection or, at least, a more rapid recovery from disease (Wraith, et al. 1987). Also, the immunity provided by killed product can be relatively short-lived (Ben-Yehuda, et al. 1993;

Couch, et al. 1996). The potential advantages of a recombinant vaccine are that it may express protective immunogens against even the most dangerous of viruses in a safe vector, provide both humoral and cell-mediated immunity, and extend the duration of that protection beyond the time provided by an inactivated product.

Vaccinia virus Vaccinia virus is a member of the Orthopoxviruses. It has made unparalleled contributions to the field of immunology. Although the virus is believed to have been initially isolated from cows around 1800 for use as an immunizing agent for smallpox, some lots of vaccine were reported to have been prepared from pox lesions in horses, or 9 mixed with smallpox virus (Moss 1991; Taylor 1993). The passage of time has obscured its origin and culture history (Baxby 1981), but one significant fact is clear: through its use, a horrifying and frequently fatal disease was eradicated. Various strains of vaccinia virus were used as Jennerian vaccines to protect humans against smallpox, with the end result of successfiilly achieving eradication of the disease in 1979 (Fenner, et al. 1988).

Vaccinia virus is considered to be a viral vector of choice because of its long history of use. Moreover, it has a wide host range, lending it to use in all mammalian species. It is environmentally durable, not requiring cold storage to retain viability, a valuable feature for usage in remote areas of the world. It is relatively nonpathogenic, with gene sequences that are well characterized. There is, in fact, a strain for which the sequence of the entire genome has been determined. The Copenhagen strain was found to have a genome of 192 kilobasepairs, with 198 open reading frames (ORF) of at least 60 amino acids, closely-spaced genes, and no introns. Other interesting characteristics and homologies included tandem and inverted terminal repeats, hairpin loops, conserved motifs for nucleotide binding sites, leucine zippers, and zinc fingers (Goebel, et al. 1990; Traktman 1990).

Vaccinia vims entry into the cell seems to be through pH-independent fiision with the plasma membrane; following penetration, there are at least two discrete stages of uncoating

(Buller & Palumbo 1991; Moss 1990). The first stage occurs at the time of internalization, consisting of the release of the phospholipid and roughly half of the viral protein; this stage may correspond to the activation of early transcription. The second stage of uncoating is characterized by the viral genome becoming sensitive to DNase I and, some suspect, corresponds to the decline of early gene transcription. One significant fact is that the vaccinia 10

\'irion contains everything it needs to begin transcription upon entry into the cell (Duller &

Palumbo 1991; Moss 1990).

Vaccinia virus has selectable non-essential genes that facilitate foreign gene insertion and, with a relatively large virus of roughly 200 kb, it is estimated that it can express up to 30 kb of inserted DNA. It is known to be a strong inducer of humoral and cell-mediated immunity. Vaccinia virus has been used as an experimental vector in several species and in accidental human exposure (Jones, et al. 1986), with considerable documentation of its ability to induce immunity to several diseases. The first reports of its use as a vector for the delivery of immunogens surfaced in the early 1980's. By 1990, there were numerous reports in the literature regarding its successful expression of the influenza virus hemagglutinin (Panicali, et al. 1983; Smith, et al. 1986) or nucleoprotein (Smith, et al. 1986), the hepatitis B virus surface antigen (Paoletti, et al. 1984), the glycoprotein D fi-om herpes simplex virus (Cremer, et al. 1985; Paoletti, et al. 1984), the rabies virus glycoprotein (Blancou, et al. 1986), envelope a glycoprotein of bovine leukaemia vims (Ohishi, et al. 1988), and other proteins.

Several studies demonstrated its immunogenicity in humans (Cooney, et al. 1993; Jones, et al. 1986). Insertion of more than one gene was also accomplished, such as one recombinant expressing both the influenza virus hemagglutinin and the herpes simplex virus thymidine kinase gene (Coupar, et al. 1988), or another expressing the hepatitis B vims surface antigen, the herpes simplex vims glycoprotein D, and the influenza vims hemagglutinin (Perkus, et al.

1985). This raised the possibility of a single vector expressing immunogens fi"om multiple pathogens, thereby providing protection against all of them with a single vaccine.

However, as promising as vaccinia vims seemed to be for certain applications, it had 11 several drawbacks which limited its potential for general use. During the smallpox eradication effort, it was known to cause unwelcome side effects. Apart from the localized irritation induced by the inoculation, there were more severe complications among immunocompromised persons. This group consists of those with AIDS and other immune system disorders; those undergoing steroidal or cancer therapy; the elderly, the alcoholic, and the debilitated; and pregnant women. It was estimated that one in every million vaccinations resulted in death (Duller & Palumbo 1992; Fenner, et al. 1988). For this reason, following the end of smallpox vaccination, vaccinia virus was no longer used clinically in the U. S. or the rest of the world. Once well-exposed to this virus, following decades of little to no use, the general population has now become quite susceptible, as people less than 30 years of age have no immunity. Unfortunately, this risk of human exposure greatly curtails the usefulness of vaccinia as a vector.

Attenuated or engineered vaccinia viruses To circumvent the problems associated with vaccinia virus, while retaining its advantages, researchers investigated strains of vaccinia virus that were attenuated, either by nature or design. Strains that were only able to produce small plaques in cell culture (Rodriguez & Esteban 1989), that were thymidine kinase negative (BuUer, et al. 1985; Taylor, et al. 1991), or that were temperature-sensitive mutants (Drillien & Spehner 1983) were evaluated. Some strains were genetically engineered to remove genes associated with virulence (Panicali, et al. 1983; Paoletti, et al. 1984;

Tartaglia, et al. 1992). From these and other studies, several strains emerged that were quite promising. One was the NYVAC strain (Tartaglia, et al. 1992), a modified form of the Copenhagen strain of vaccinia virus, which had been genetically engineered so that it no longer was able to produce infectious virions. Another strain was the modified vaccinia

Ankara (MVA) strain which also was replication incompetent. Its attenuation was the result of about 570 passages in chick embryo fibroblasts, an unnatural host, and not the result of direct engineering. Nonetheless, its genome had been investigated and the areas of deletion determined (Meyer, et al. 1991). Although no studies have yet directly compared these two host-restricted poxviruses, the most critical difference between the two would appear to be that the NYVAC strain has a deficiency in replication at an earlier stage than the MVA strain.

This would mean less protein production with the NY\''AC strain than with MVA, which allows for both early and late gene expression. The implication for immunization would be that of less protection following use of the NYVAC vector containing foreign gene inserts.

Nonetheless, its successful use as a vaccine vector has been reported in the literature, where expression of foreign antigens provided protection against pseudorabies virus and Japanese encephalitis virus in swine (Brockmeier, et al. 1993; Konishi, et al. 1992).

Alternate poxvirus vectors In addition to pursuing attenuated or engineered strains of vaccinia virus, researchers pondered the benefits of using strains of host-specific poxviruses as vectors to deliver genes in unnatural hosts. It seemed reasonable to assume that, if a modified vaccinia virus so impaired that it cannot replicate is still able to induce immunity, then an intact host-restricted poxvirus should be able to do the same. Several of these have proven to be successful, specifically the avian poxviruses, one of which is currently being used as a licensed veterinary product. As mentioned above, canarypox virus 13 has been used as a vector for a canine distemper vaccine. Canarypox recombinants have also been developed that express the measles virus fusion and hemagglutinin glycoproteins

(Taylor, etal. 1992), the rabies glycoprotein (Cadoz, etal. 1992), and feline leukemia virus env and gag proteins (Tartaglia, et al 1993).

Another approach has been to use host-restricted poxvirus vectors in their natural hosts, as vectors to deliver immunogens for protection against other diseases. Examples of this strategy include the licensed fowlpox virus vaccines that serve as vectors for avian influenza virus (Webster, et al. 1991; Webster, et al. 1996) and Newcastle disease virus

(Taylor, et al. 1996) genes. The advantage in using this as a vector for poultry is that poultry are also vaccinated against fowlpox virus, so this vaccine construct provides several useful purposes. Another such candidate vector is swinepox virus. It has been described in the literature as successfully expressing genes of pseudorabies (van der Leek, et al. 1994). One concern regarding use of swinepox virus is that it can, on occasion, be pathogenic in its host.

Also, given that it can and will replicate in its host, the potential exists for uncontrolled spread of a swinepox recombinant among domestic and feral swine.

Host range restriction of the MVA strain of vaccinia virus The N'lVA strain of vaccinia virus has been well characterized (Altenburger, et al. 1989; Antoine, et al. 1996;

Bender, et al. 1996; Carroll & Moss 1997; Carroll, et al. 1997; Hirsch, et al. 1996; Mayr, et al. 1975; Mayr, etal. 1978; Meyer, etal. 1991; Scheiflinger, etal. 1996; Sutter & Moss

1992; Sutter, et al. 1994; Wyatt, et al. 1996). It was originally developed from the vaccinia virus Ankara strain as a safe alternative for smallpox vaccination, and has been used without 14 significant side-effects in over 120,000 people, including young children and the elderly, for immunization against smallpox. After approximately 570 passages in primary chick embryo fibroblasts (CEF), it lost its ability to replicate or at least replicate well in numerous manmialian cell lines. It contains six major deletions that prevent virus assembly in mammalian cells, however, leaving gene expression, both early and late, relatively unimpaired. The exact nature of this host restriction is not really understood. Thus far, four orthopoxvirus host-range genes have been identified. These are the CHOhr (Gillard, et al.

1985), C7L (Oguiura, et al. 1993; Perkus, et al. 1990), KIL (Perkus, et al. 1990), and E3L

(Beattie, et al. 1996; Chang, et al. 1995) genes. Of these, only the function of the E3L gene, which expresses an RNA binding protein (Chang, et al. 1992), is known. Regarding the others, the CHOhr gene is required for vaccinia to replicate in Chinese hamster ovary cells; the CHOhr, KIL, or C7L gene is required in human MRC-5 and porcine kidney PK-15 cells;

KIL or CHOhr is required in rabbit kidney RK13 cells; and E3L is required in Vero and

HeLa cells. Compared to its parental strain, MVA has deletions that consist of about 15%

(30,000 base pairs) of its former genome, including most of the KIL gene (Altenburger, et al.

1989; Meyer, et al. 1991). Interestingly, in one study, replacement of the KIL gene in MVA removed only the host restriction in RK13 cells (Carroll & Moss 1997). This suggests that there are multiple, cumulative genetic defects in MVA replication. If so, as seems likely, the probability of spontaneous reversion to a wildtype host range is quite low, which increases the safety of MVA as a vaccine vector.

Most host range mutants do not have as broad a host range restriction as that of

MVA. One study (Meyer, et al. 1991) described the lack of viral multiplication in many 15 different cell lines, including human cervix (HeLa), lung (MRC 5), colon (HRT 18), and larynx (Hep-2); monkey kidney (Vero); rabbit kidney (RK13); equine dermal (E-derm); bovine lung (BEL) and kidney (MDBK); canine kidney (MDCK); and mouse (DBT) cell lines. There was evidence of some replication, reduced about 100-fold, in only two cell lines tested, monkey kidney (MAI 04) and chick fibroblast (LSCC-H-32). In a recent study supporting these findings (Carroll & Moss 1997), it was determined that MVA could not replicate (< 1-fold increase) in rabbit cornea (SIRC), rabbit skin (RAB-9), rabbit kidney

(RK13), pig kidney (PK-15), human cervix (HeLa), human kidney (SW 839, 293), rhesus monkey kidney (FRhK-4), Chinese hamster ovary (CHO), and Chinese hamster lung (CHL) cell lines. A few cell Unes were found to be semi-permissive (1- to 25-fold increase) to MVA growth, namely, Afiican green monkey kidney (BS-C-1, CV-1), and canine kidney (MDCK) cell lines. Only two cell lines, both fibroblastic in morphology, were permissive (> 25-fold increase). These were a quail embryo (QT35) and a Syrian hamster kidney (BHK-21) cell line, neither of which achieved the levels of virus replication demonstrated by MVA in CEF cells.

The host restriction of MVA, although demonstrated in many mammalian cell lines

(Carroll & Moss 1997; Meyer, et al. 1991), has only been explored in one swine cell line. If the inability to assemble infectious MVA virions extends to various swine cells, the risk of spread to nonvaccinated individuals or the environment is minimal. Such host-restriction, coupled with a large double-stranded DNA genome capable of accomodating 25 kb of foreign DNA, make MVA an attractive candidate for SIV vaccine. Moreover, MVA has been used successfiilly in the past to vector influenza virus structural proteins in mice 16

(Bender, etaL 1996; Sutter, etaL 1994), parainfluenza virus 3 proteins in cotton rats (Wyatt,

et al. 1996), and simian immunodeficiency virus env-gag-pol in macaques (Hirsch, et al.

1996). In addition, it was able to protect against and provide therapy for pulmonary

metastases in mice (Carroll, et al. 1997).

MVA is also a practical choice for laboratories interested in developing recombinant

vaccines. Since 1991, it has been recommended that U. S. researchers working with standard

vaccinia viruses receive smallpox vaccinations every 10 years and that their work be carried out in biosafety level 2 containment (Katz & Broome 1991). Prior to 1991, restrictions were even more stringent. Underscoring the avirulence of MVA, the National Institutes of Health intramural biosafety committee in 1997 removed the vaccination requirement for work with

MVA and changed its status to biosafety level 1 contairmient (Carroll &. Moss 1997).

Nomenclature of Orthomvxoviridae Swine influenza virus (SIV) is a member of the family Orthomyxoviridae. Influenza viruses are grouped into types A, B, and C on the basis of their nucleoprotein and matrix protein similarities (Lamb & Krug 1996; Murphy &

Webster 1996). Type A contains, by far, the most pathogenic strains and can infect a wide variety of mammalian and avian species. In 1980, the World Health Organization revised the system of nomenclature, in the face of increasing awareness of antigenic diversity among isolated strains, to provide the information on type, host if nonhuman, geographical location, strain number, and, for type A viruses, designation of antigenic specificity (subtype) of the surface antigens, hemagglutinin (here, H; also abbreviated HA in other contexts) and neuraminadase (N) (Schild, et al. 1980; WHO 1980). By convention, the subtypes are given 17 in parentheses after the rest of the information is provided. An example of this v/ould be

A/Sw/IN/1726/88 (HlNl), denoting a type A swine strain, # 1726 from Indiana, subtype

HlNl, isolated in 1988.

Subtypes of influenza Type A viruses Currently, there are 15 H and 9 N subtypes, reflecting the ability of these antigens to be modified by environmental or immunological selection (Rohm, et al. 1996). The RNA genome undergoes frequent point mutations, on occasion resulting in evolutionary advantage, in a process knovm as antigenic drift. But, the segmented genome itself can reassort into new combinations of H and N in the presence of another subtype. This process is referred to as antigenic shifts and can result in the introduction of new subtypes into a susceptible population. Interspecies transmission, mixed infections, and gene reassortment are thought to be responsible for the emergence of new human pandemic strains (Webster, et al. 1995).

The current consensus is that aquatic birds are the reservoirs for all subtypes of influenza A viruses (Webster 1998). Waterfowl are enterically infected wth virus, which is excreted in high quantities into the waterways that the birds frequent. This provides an efficient means for spreading virus to other wild and domestic animals. Avian-origin viruses are believed to be responsible for outbreaks of influenza in poultry (Horimoto, et al. 1995) and various mammals, e.g. whales (Hinshaw, et al. 1986), seals (Geraci, et al. 1982), and pigs (Scholtissek, et al. 1983). Once virus is spread to humans, pigs, or horses, the route of infection is primarily respiratory.

Phylogenetic analysis has indicated the strong probability that all mammalian influenza 18 viruses are derived from an avian ancestor. Strains of ail subtypes, from all species and geographic regions, have been sequenced and analyzed. From examination of highly conserved sequences in the genome, it would appear that there are five lineages: gull, swine, human, 'ancient' equine (not seen since over 15 years ago), and recent equine (Webster

1998). Of these, swine and human strains have apparently evolved from a conmion ancestor.

There also appear to be sublineages, due to longitudinal migration patterns of waterfowl, such that Europe and Asia have strains that can be distinguished from those isolated in North and South America. The inability to restrict the movement of a diverse and widespread reservoir species requires that efforts against influenza be directed towards prevention and control.

Influenza virus genome The genome consists of eight (types A and B), seven

(type C and Dhori virus), or sbc (Thogoto virus) molecules of single-stranded, negative-sense, linear RNA for a total of 10-13.6 kb (Murphy 1996). Virions are 80-120 nm in diameter, pleomorphic, and enveloped, with large peplomers composed of HA and N proteins projecting from the surface. The type A and B viruses have, for structural proteins, three polymerase proteins (PA PBl, and PB2), a nucleocapsid protein (NP), a hemagglutinin

(HA), a neuraminidase (N), and a nonglycosylated matrix protein (M or Mi). There are also two nonstructural proteins (NSl, NS2) and an ion channel (Mj). The eight segments of

RNA, ranging from 890 to 2,341 nucleotides, interact with NP to form a ribonucleoprotein

(RNP) which associates with the transcriptase complex consisting of PBl, PB2, and PA, to form the nucleocapsid (Lamb & Krug 1996). Swine influenza virus Swine influenza (SI) is a common disease of pigs caused

by type A influenza viruses, mostly of the HlNl antigenic subtype (Easterday & Hinshaw

1992). It was first described in 1918, at the time of the great influenza pandemic responsible

for the death of 20 million people worldwide, but the virus was not isolated until 1930

(Easterday & Hinshaw 1992; Shope 1931). Presently, it occurs throughout much of the

world wherever pigs are found. The disease is characterized by pyrexia, anorexia, dyspnea,

coughing, sneezing, depression, huddling, and pneumonia. In acutely affected herds, it

exhibits 100% morbidity, although mortality is low. Without depopulation, SI is likely to

continue in the herd with episodic occurrences of respiratory disease and reproductive

problems throughout the year, not just seasonally as was once thought (Nakamura, et al.

1972). It is currently believed that convalescing pigs may serve as carriers, or as a reservoir of SIV, between epizootics.

With an estimated one-third of pigs in the US affected (Chambers, et al. 1991;

Easterday Hinshaw 1992; Hinshaw, etal. 1978; Pirtle, et al. 1976; Woods 1975), the economic impact of SIV on the swine industry is substantial. Even though SFV is not usually extremely virulent, pigs can become very lethargic, pyrexic, and anorexic, generally up to about 5 days. With other pathogenic agents present in a herd, such as porcine reproductive and respiratory syndrome virus (PRRS), dual synergistic infections can incapacitate a herd.

SIV prophylaxis Presently, there is one killed product, but no live SIV vaccines are commercially available. No strain has yet been found to be both efficacious and sufficiently attenuated to serve as a safe live virus vaccine in pigs. In addition, there are 20 ongoing public health concerns regarding the transmission of SFV to humans, with the most recent fatality reported in 1994 (Wentworth, et al. 1994). A swine population that is vaccinated against SIV would protect agricultural workers from exposure to zoonotic SIV strains. Thus, for several reasons, evaluation of the safety and efiBcacy of a recombinant SIV vaccine is desirable. As the antigenic characteristics of the dominant SIV strains are very stable, without the marked drift observed in human strains, SFV would seem an excellent candidate for control by vaccination. For other subtypes of SIV and newly emerging variants, additional genes may be inserted to broaden the recombinant's range of immunogenicity.

Mode of infection Infection occurs through inhalation of aerosolized virus, resulting in attachment of virus to respiratory epithelium through binding of the distal tip of the HA surface protein to sialic acid receptors on the susceptible cell. Virus is endocytosed into endosomes which fuse with lysosomes. The resulting acidic pH induces a conformational change in HA which exposes a hydrophobic region of HA that can then fuse with the endolysosome membrane, releasing ribonucleoprotein into the host cell cytoplasm

(Bullough, et al. 1994; Lamb & BCrug 1996). Detailed analysis of HA has determined that it is, in fact, a trimer of identical subunits, each of which has an HAi (328 residues) and an HA,

(221 residues) linked by a disulphide bond. The precursor HAo must be cleaved into HAj and

HA, to allow conformational change to the low pH form required for infectivity. If the connecting peptide has the sequence R-X-BC/R-R (where R is the basic amino acid arginine, X is a nonbasic amino acid, and K is the basic amino acid lysine), cleavage of HA occurs in the 21 trcms Golgi apparatus by the endogenous protease fiirin. Certain strains of HA having

multiple basic amino acids adjacent to the cleavage site have been associated with high

pathogenicity in avian strains (Lamb & BCrug 1996; Senne, et al. 1996).

Clinical pathology of SIV The pathology induced by SFV occurs throughout the respiratory tract and consists of acute inflammation, edema, and necrosis (Murphy &

Webster 1996). More severe complications, in the form of interstitial pneumonia, thickening of alveolar walls, hyperemia, thrombosis, hemorrhage, and necrosis, can occur. Lung lesions tend to be bilaterally distributed, predominantly in the cranial and middle lobes. In pigs that recover, resolution of lesions may take up to a month.

The internationally recognized method for evaluation of pathogenicity of avian influenza virus (AIV) isolates is by experimental inoculation of chickens. An isolate that causes the death of at least six of eight (75%) inoculated 6-week-old susceptible chickens is considered "highly pathogenic" (HP) (Senne, et al. 1996). Generally, the H5 and H7 isolates are not HP, producing mild, localized infections of the respiratory and intestinal tracts.

However, systemic infection can occur, producing the acute disease that often results in death. Presently, there is no correlate nomenclature for SIV because, traditionally, SFV has not incited acute, fulminating, systemic disease. Typically, SFV virulence might be better measured in terms of percent of pneumonic pigs among those exposed. However, in recent years, there have been periodic occurrences of'atypical' SIV outbreaks. This has led to speculation that the relatively stable antigenic profile of SIV, at least in the U. S., may be changing. 22

Atypical strains of SIV Reports have appeared in the literature since 1992,

indicating the occurrence of SIV either associated with unusual signs or exhibiting more

virulence than expected. First, there was a report from Quebec regarding an HlNl

variant producing proliferative and necrotizing pneumonia in pigs, with some signs very similar to those of the PRRS virus (Dea S., et al. 1992). It may be that this was an instance of undetected dual infection, but genetic analysis did reveal more point mutations and diversity than generally seen in North American SIV isolates (Rekik, et al. 1994). Other cases arose, as well, indicating novel strains of SIV. One such isolate from a severely affected herd was designated A/Sw/Nebraska/1/92 (Olsen, et al. 1993). It induced persistent, high fevers (up to 42°C) but not much respiratory disease. Given the high degree of conserved sequences in U. S. classical SIV strains, it was surprising that the most closely related reference HlNl strain had only 94% identity at the nucleotide level and 96% at the amino acid level to this SIV isolate. Nonetheless, it was closest genetically to classic HINl

SIV than to avian or human HlNl viruses. In England, an HlNl strain antigenically distinguishable from classic SIV and European avian virus-like HlNl viruses caused a sudden increase in SIV cases, but still exhibited the usual clinical signs of coughing, sneezing, and anorexia (Brown, et al. 1993). However, upon experimental infection, this strain produced a more severe interstitial pneumonia and hemorrhagic l^nnph nodes.

The significance of the genetic diversity represented by these strains is as yet undetermined. It may be that there are no 'atypical' SIV strains, merely a greater degree of potential antigenic diversity among field strains than previously noticed. Interspecies transmission and reassortment among influenza vimses Subtype

HlNl influenza viruses have been continuously circulating in U. S. pigs for over 60 years. It was believed that the great pandemic of "Spanish flu" in 1918/19, the worst in history, killing at least 20 million people worldwide, was either caused by a swine virus or by a human strain that entered the pig population at that time (Kaplan & Webster 1977). In 1997, RNA from a person who died during that pandemic was extracted from formalin-fixed, paraSin-embedded tissue and sequenced (Taubenberger, et al. 1997). All sequences determined were very similar to those of classic HlNl SFV, suggesting that human and swine strains share a common avian ancestor, existing some time before 1918. The first isolation of SIV did not occur until 1930, however, and the 'classical' HlNl swine virus recovered

(A/Swine/Iowa/15/30) is still much like the majority of SFV found circulating in U. S. pigs today. By sequence analysis of the NP genes from various species over time, it was recently determined that, at the nucleotide level, the classical SFV NP is human virus-like, but at the amino acid level is avian virus-like. Some suggested this implied selection pressure on the

SIV NP, acquired early from humans, to revert to avian sequences perhaps because of a concomitant reassortment with an avian virus. Presumably, the presence of other avian- origin genome segments would preferentially select for a more avian virus-like NP, given

NP's critical role in replication (Gammelin, et al. 1989). But this does not appear to be the case. Sequence analysis of additional gene segments of classical SFV reveal that most of them, as with NP, are more closely related to avian strains, suggesting they ail evolved simultaneously (Schultz, et al. 1991). Since it is known the human virus NP is under strong pressure to change (Gammelin, et al. 1990; Gorman, et al. 1991), current speculation is that 24 around 1920 a human HlNl strain was transferred into U. S. pigs, thereby releasing that strong selective pressure, which in turn allowed the (classical SIV) genes to slowly start to evolve back to the original, optimal avian sequences from which mammalian influenza strains originated.

This phenomenon of slow antigenic drift should not be confused, however, with the appearance of avian virus-like strains in Europe. Northern Europe saw its first isolate of SIV in 1978/79, and although an HlNl virus, its HI was similar to the avian HI but distinct from both human and swine HI (Hinshaw, etal. 1984; Scholtissek, etal. 1983). Since then, there have been instances where an avian virus has been able to cross species and infect the pig population, as with SwGer/81, and cases where reassortment between avian and classical SIV has occurred, as with SwHK/82 (Schultz, et al. 1991). In addition, there is a human virus- like H3N2 subtype that has been isolated on occasion in European pigs since 1980, perhaps as a result of the 1968 antigenic shift (Kundin 1970; Shortridge, et al. 1977) and an ability of the H3N2 to persist in pigs even when not circulating in the human population (Done &

Brown 1997). It is interesting that a serological survey conducted in 1988-1989 in U. S. pigs found evidence of H3 viruses antigenically similar to the then-current human H3 strains, at about 1.1% average prevalence (Chambers, etal. 1991). In addition, a serological survey during 1976-1977 detected an incidence of 1.4% for H3N2 infections (Hinshaw, et al. 1978).

Moreover, in that study, isolation from one herd of a virus antigenically similar to a human

H3N2 strain was reported. However, complete sequencing to determine whether the isolate was of human origin was not performed. So, no H3 human strain has been confirmed as present in U. S. pigs. Very recently, the National Veterinary Services Laboratories identified 25 an influenza virus subtype H3N2 isolate from a swine breeding herd in eastern North Carolina

(interlaboratorj' communication). Abortion and typical signs of influenza were reported in the herd, with mortality at approximately 10%. Studies are currently underway to determine species of origin and other characteristics of the virus.

So, at the very least, there are three HA subtypes circulating in pigs at present, classic

SrV HI, avian virus-like HI, and human virus-like H3. They have been found in various permutations of SIV gene segments. One, an H3N1 strain, appeared to be a combination of the classic SIV and the human virus-like H3N2 found in swine (Done & Brown 1997).

Another, an H1N2 isolate believed to be from a human HlNl and the swine-adapted H3N2, caused clinical disease in pigs (Brown, et al. 1995). Still another represented a reassortment between human and avian strains in symptomatic Italian pigs, providing the first proof that pigs can act as 'mixing vessels' for human and avian viruses (Castrucci, et al. 1993). The critical role that pigs can play in pandemics was underscored with the discovery that children in the Netherlands were sick from avian-human influenza virus generated in pigs, transmitted pig-to-person, and person-to-person (Claas, et al. 1994). Normally, avian strains do not replicate in humans, and human strains do not replicate in birds. This is a flinction of their specific sialyloligosaccharide receptors on the surface of epithelial ceils of the upper respiratory tract. In a previous study, it was determined that, of 38 avian influenza strains, 31 were successfully transmitted to swine (Kida, et al. 1994). Every HA subtype (of 14) had at least one strain that grew as well as a swine or human virus. Since then, it has been determined that pigs, in fact, have both avian- and human-specific viral receptors present in their upper respiratory tract and that some avian strains, with continued replication, acquire 26 the ability to recognize human receptors as they become swine-adapted (Ito, et al. 1998).

Taken together, these data delineate the danger to humans of a pig population unprotected against influenza. If an avian virus with a non-human-type HA is introduced into pigs, then reassorts with a human strain, a pandemic among complete susceptibles would occur. Although direct interspecies spread from bird to human can happen, as was seen with the 1997 Hong Kong H5NI cases (Claas, etal. 1998a; Claas, etal. 1998b; Suarez, et al.

1998; Yuen, etal. 1998), the virus under those circumstances may not readily adapt to its new host and relatively few may be affected. The dangers may be greater in the former scenario.

In summary, interspecies transmission is known to occur and the pig can be infected by either avian or human strains and can serve as a 'mixing vessel', wherein gene segments from different strains can reassert to produce new viruses. Clearly, a safe, live vaccine vector with capacity to express multiple genes, that could be given to induce primary immune response or to boost immunity, would be helpful.

Immunitv to influenza viruses There are numerous factors that influence the ability to withstand influenza virus infection. One is the extent of antigenic variation, through drift or shift (Askonas, et al. 1982; Yetter, et al. 1980). The more novel the antigenic profile, the more susceptible a population of individuals will be. In addition, the immune status of the host plays a large role in protection against influenza (Yetter, et al. 1980). The very old and the very young, with reduced capacity for generating humoral and cell-mediated immunity, are especially vulnerable to influenza infection (Ben-Yehuda, et al. 1993). But even those 27 with a robust immune system are subject to variables which affect the magnitude of the immune response generated. These include the history of prior exposure and infection, the time since last vaccination, and the degree of antigenic similarity between the vaccine and the circulating strain of influenza virus (Couch, et al. 1996).

Other factors that influence variation in severity and outcome of exposure to influenza virus are the site of initial infection and heterotypic immunity (Yetter, et al. 1980). It has been determined in mice that an infection initiated throughout the respiratory tract can lead readily to a fatal viral pneumonia, whereas an infection initiated in the nares will seldom do so. Moreover, once exposed and recovered, mice display a decrease in viral shedding, reduced pulmonary infection, and enhanced recovery when exposed to heterotypic strains of influenza (Liang, et al. 1994; Yetter, etal. 1980).

This ability to recover more rapidly from virus infection, but not prevent it, has been attributed to the action of cross-reactive type-specific cytotoxic T lymphocytes generated by

NP priming (Wraith, et al. 1987). Immunity to influenza virus requires antibodies to the globular region of the HA molecule. When these are present at the site of virus exposure, in sufBcient quantity, the virus is neutralized and infection prevented (Andrew & Coupar 1988;

Couch & Kasel 1983; Smith, et al. 1983). But even if the HA is too different for humoral recognition and infection is initiated, other components of the immune system come into play.

Interestingly, one study (Liang, et al. 1994) found that the mechanisms are partially different between the upper and lower respiratory tract. In the nose, both CD4+ and CD8+ T cells were almost entirely responsible for a short-lived heterotypic immunity. In the lungs, the short-lived effects of CD8+, but not CD4+, T cells enhanced recovery, but only partly. 28

Some other undetermined mechanism was responsible for a more persistent heterotypic protection. Depletion of NK cells did not have an effect on protection in either the lung or nose.

All evidence to date indicates an optimal influenza vaccine should induce both neutralizing antibodies and cytolytic T cells to eliminate free virus and infected cells. One phenomenon that may affect the performance of vaccines in immunization against influenza is that of'original antigenic sin', a term first used to describe the antibody response to influenza virus. This was first observed in individuals after an initial infection then reinfection, or vaccination, with a new strain of virus. The response was found to be that of boosting antibody titers specific to the original infecting strain which only weakly cross-reacted to the new strain (de St. Groth & Webster 1966; Francis 1953). Recently, the concept of original antigenic sin has been extended to C5rtotoxic T Ijnnphocytes (CTL) (Klenerman &

Zinkemagel 1998). Following infection of mice with an immunodominant strain of lymphocytic choriomeningitis virus, the mice had only a slight cross-reactive response when challenged with a mutant strain, reacting primarily to the first virus. If the mutant strain was given first, this response was not observed. Instead, CTL responses cross-reacted equally well to the mutant and immunodominant strains. Although the mechanism is not yet understood, reactivation of strong memory CTL overcomes the slower response of CTL precursors (McMichael 1998). How closely related, yet disparate, viral strains have to be for this phenomenon to occur, at either the humoral or cell-mediated level, has yet to be determined. But the occurrence of original antigenic sin speaks to the need for vaccination prior to any infection that will strongly prime the immune system. 29

Selection of an SIV vaccine SIV strains in the U. S. have exhibited little antigenic variation over time, with one dominant antigenic type responsible for most disease outbreaks. This phenomenon has been attributed to the lack of selection pressure provided by the relatively short lifespan of swine and the continual presence of a susceptible population, i.e. young pigs without antibodies (Easterday & Hinshaw 1992; Luoh, et al.

1992). It has previously been demonstrated that monoclonal antibodies generated against the human A/PR/8/34 (PR/8) strain are cross-reactive with antigenic sites on A/SwAne/Iowa/31

(IA/31), a classic example of the dominant antigenic type of SIV in pigs (Gerhard, et al.

1981). It has also been shown that the HA of A/Sw/In/1726/88 (IN/88) has three antigenic sites that correspond to the three sites on the HA of PR/8, located in the loop and distal tip of the protein (Luoh, et al. 1992). These findings suggest that some degree of cross-reactivity will occur between antibodies generated against the PR/8 strain and various swine influenza virus isolates.

Consequently, in the first study of this thesis, the ability of a recombinant iVTVA vector containing influenza virus HA and NP genes fi-om the PR/'8 strain (MVA/PR8) to provide immunity to SIV will be evaluated. During a three-month sabbatical in 1992 at the

National Institutes of Health, Bethesda, MD, I participated in the creation of this recombinant, cloning the HA and NP genes into intermediate and transfection vectors, then preparing plasmid DNA for transfection into MVA (Sutter, et al. 1994). In the second study of this thesis, an MVA recombinant expressing SIV HA and NP genes will be constructed, then compared to MVA/PR8 for efiBcacy in protecting pigs against SIV challenge. The strain providing the genes will be a field strain isolated in 1992 by the Veterinary Diagnostic 30

Laboratory at Iowa State University, Ames, lA. It has been designated A/Sw/IA/40776/92

(ISU) and is believed to be of the classic HlNl variety, on the basis of the clinical disease it produces and the site of its isolation. The HA and NP genes of this strain will be sequenced in the course of these studies and compared to those of a standard HlNl SFV to test this supposition.

MVA vectors used The first of the two MVA recombinants used in these studies contains the HA and NP genes firom PR8 in opposite orientation and under the control of two optimized synthetic early/late promoters, improved over wild type vaccinia virus promoters, following base-by-base analysis of performance (Davison & Moss 1989a;

Davison & Moss 1989b). It has been shown to provide protection to mice against a lethal challenge fi"om PR/8 (Sutter, et al. 1994). The second MVA-vectored construct expresses the HA and NP genes firom the field strain ISU, also in opposite orientation and using the same strong synthetic promoters as the PR8 vector.

The significance of HA and NP as immunogens The use of the HA and NP gene inserts was not serendipitous. Previous studies have demonstrated that IL\ can induce virus-neutralizing antibody (Andrew, et al. 1987; Ben-Yehuda, et al. 1993; Smith, et al.

1983); however, NP induces only non-neutralizing antibody (Wraith, et al. 1987). NP, however, is the major target antigen recognized by type-specific cross-reactive cytotoxic T- lymphocytes (CTL) (Townsend & Skehel 1984; Yewdell, et al. 1985) whereas HA only stimulates subtype-specific CTL (Andrew, et al. 1986; Bennink, et al. 1984). It is known that 31 antibody is important for protection against infection (Webster, et al. 1991); CTL response is responsible for recovery from infection (Wraith, et al. 1987). Since most of the protective humoral response following exposure to influenza virus is generated against the five hypervariable antigenic sites of HA, antibodies to this protein provide immunity to influenza viruses with a sufficiently similar HA (Andrew, et al. 1987).

Route of inoculation The optimal route of vaccine delivery for a respiratory pathogen, such as influenza virus, has been a matter of some discussion. One study (Small, et al. 1985) found that intranasal (IN) administration of a W recombinant expressing HA, in mice protected both lung and nose against homologous challenge, but scarification protected only lung. Another group (Meitin, et al. 1991) vaccinated mice either intraperitoneally (IP) with a killed PR8 vaccine, stimulating high serum IgG, or IN with a W recombinant containing HA,, inducing nasal IgA titers. They found that the lungs but not noses in the IP group and the noses but not lungs in the IN group were fixlly protected against challenge. By reversing protection in mice recovered from infection with influenza virus, using anti-IgA antiserum, it has since been shown that IgA is the primary and perhaps the only factor in nasal immunity to influenza virus in mice (Renegar & Small Jr. 1991). To maximize both IgA and

IgG, several studies have tried to stimulate the 'common mucosal immune system' (Yetter, et al. 1980) in mice by intragastric or intrajejunal administration of a W or MVA recombinant expressing influenza virus genes (Bender, et al. 1996; Meitin, et al. 1994). They, in fact, have been successfiil in attaining mucosal IgA, serum IgG, and CTL activity. Consequently, the use of orally-administered enteric-coated capsules was considered for these present 32 studies, but then deemed not a viable option for weanling pigs, though they hold great promise for human vaccination programs.

Inhibition of immune response to secondary vaccination The advent of recombinant vaccines capable of expressing numerous foreign inserts has raised the question of whether frequent use of such constructs to deliver different antigens might be inherently nonproductive (Etiinger & Altenburger 1991; Flexner, etal. 1988; Rooney, et al. 1988).

Would, in fact, preexisting immunity to the virus adversely affect efficacy? It has been thought by some that the immune response generated against a vector would limit its replication and prevent adequate expression of the inserts. This, in turn, would inhibit the induction of protective immunity against the heterologous proteins. In support of this view, one study (Chelyapov, et al. 1988) demonstrated the strength of the immune response to W.

It showed that antibodies were generated against most of the W structural proteins, including those proteins located internally within the virus, in both rabbits and humans, that these antibodies in humans were preserved over many years, and that insertion of foreign genes did not affect this pattern. Additional reports indicated that preexisting immunity to

W resulted in reduced titers of antibody to the foreign protein (Cooney, et al. 1991;

Rooney, et al. 1988), decreased protection (Rooney, et al. 1988), and reduced, transient T- cell response (Cooney, et al. 1991). Another study (Kundig, et al. 1993) demonstrated that vaccination with one W recombinant resulted in long term suppression of the humoral response in mice to a second W recombinant's gene product.

On the other hand, there have been reports of boosting antibody titers in animals 33 given the same recombinant virus, or of induction of immunity against a foreign antigen in animals vaccinated with W subsequently inoculated with a recombinant W expressing that antigen (Etiinger &. Altenburger 1991; Jones, et al. 1986; Perkus, etal. 1985; Rooney, et al.

1988). Also, antibody titer to one of the heterologous proteins in a multivalent W recombinant did not prevent mice from developing immunity (Flexner, et al. 1988). Still, questions have persisted about the adequacy of the response in these cases, even if outright suppression has not occurred.

Interference by maternal antibodv Another concern has been that, even if anti-

W antibody has little effect on the antibody response to a foreign gene product, there may stiU be interference from pre-existing maternal antibody to the virus from which the protein derives. Some have noted that passively administered polyclonal antibody to a foreign protein can inhibit B-cell response to a W recombinant vaccine, but not necessarily the cytotoxic T-lymphocyte response (Gralletti, et al. 1995; Johnson, et al. 1993). In one study

(Brockmeier, et al. 1997), it was found that only one of two W recombinants expressing different pseudorabies glycoproteins protected equally well, with or without the presence of maternal antibody. This finding suggests that an appropriate choice of insert may allow induction of immunity in the presence of maternal antibody.

Proposed inoculation regimens to improve immunitv To avoid interference in secondary vaccination and improve the immune response, some have tried using diversified prime and boost regimens for vaccinations, as in priming with a VY recombinant, followed by 34 a subunit vaccine containing the expressed protein. Use of a subunit as booster has met with mixed results (Cooney, et al. 1993; Montefiorl, et ah 1992). One study found that use of a

W recombinant, boosted by a recombinant avian pox virus, resulted in improved T-cell responses, better than when using either construct alone (Hodge, et al. 1997). Another study determined that mice primed with an influenza virus recombinant and boosted with a W construct expressing the same antigen developed strong secondary antigen-specific CDS' T cell responses, but only if the vaccines were given in just that order. However, these regimens are not appropriate for achieving widespread prophylaxis and therapy on a commercial scale. From a practical standpoint, the preferred alternative would be to have just one vector, capable of being used safely and repeatedly to express numerous inserts.

Statement of the Problem

To date, no work has been presented regarding the use of any W recombinant expressing influenza virus proteins for the protection of pigs against SIV. Nor has there been any interference study on the MVA strain of W which, though replication incompetent, may express sufficient protein to inhibit immune response. These issues will be explored in the first, second, and third studies of this project.

It is possible that the HA and NP genes fi"om swine influenza virus, especially those fi"om pathogenic isolates, will provide as good or better immunity to this disease than the current inactivated vaccine. In addition, the MVA vector may prove to be a valuable delivery system for other viral immunogens. Its extensive history of use, detailed characterization, and 35 strong synthetic promoters make it a model candidate for expressing proteins of pathogenic agents. Our ability to test this system now against swine influenza helps us to evaluate its potential to prevent other diseases and, at the very least, provides additional data useful in evaluation of the entire recombinant vector concept.

References

Altenburger W, Suter C-P, AJtenburger J. 1989. Partial deletion of the human host range gene in the attenuated vaccinia virus MVA. Arch Virol 105:15-27.

Andrew ME, Coupar BEH. 1988. Efficacy of influenza haemagglutinin and nucleoprotein as protective antigens against influenza virus infection in mice. Scand J Immunol 28:81-5.

Andrew IVDE, Coupar BEH, Ada GL, Boyle DB. 1986. Cell-mediated immune response to influenza virus antigens expressed by vaccinia virus recombinants. Microb Pathog 1:443-52.

Andrew ME, Coupar BEH, Boyle DB, Ada GL. 1987. The roles of influenza virus haemagglutinin and nucleoprotein as protective antigens against influenza virus infection in mice. Scand J Immunol 25:21-8.

Antoine G, Scheiflinger F, Holzer G, Langmann T, Falkner FG, Domer F. 1996. Characterization of the vaccinia MVA hemagglutinin gene locus and its evaluation as an insertion site for foreign genes. Gene 177:43-6.

Askonas BA, McMichael AJ, Webster RG. 1982. The immune response to influenza viruses and the problem of protection against infection. Beare AS, ed. Basic and applied influenza research. Boca Raton, FL: CRC Press, p 159-88.

Baxby D. 1981. The origins of vaccinia. Jennet's Smallpox Vaccine. London: Heinemann. p 179-96.

Beattie E, Kauffman EB, Martinez H, Perkus ME, Jacobs BI, Paoletti E, Tartaglia J. 1996. Host-range restriction of vaccinia virus E3L-specific deletion mutants. Virus Genes 12:89-94. 36

Ben-Yehuda A, Ehleiter D, Hu A, Weksler ME. 1993. Recombinant vaccinia vims expressing the PR/8 influenza hemagglutinin gene overcomes the impaired immune response and increased susceptibility of old mice to influenza infection. J Infect Dis 168:352-7.

Bender BS, Rowe CA, Taylor SF, Wyatt LS, Moss B, Small Jr. PA. 1996. Oral immunization with a replication-deficient recombinant vaccinia virus protects mice against influenza. J Virol 70(9):6418-24.

Bennink JR, Yewdell JW, Smith GL, Moller C, Moss B. 1984. Recombinant vaccinia virus primes and stimulates influenza haemagglutinin-specific cytotoxic T cells. Nature 31I(5986):578-9.

Blancou J, Kieny M-P, Lathe R, Lecocq JP, Pastoret PP, .Soulebot JP, Desmettres P. 1986. Oral vaccination of the fox against rabies using a live recombinant vaccinia virus. Nature 332:373.

Brochier B, Thomas I, Bauduin B, Leveau T, Pastoret P-P, Languet B, Ghappuis G, Desmettre P, Blancou J, Artois M. 1990. Use of a vaccinia-rabies recombinant virus for the oral vaccination of foxes against rabies. Vaccine 8(2):101-4.

Brockmeier SL, Lager BCM, Mengeling WL. 1997. Vaccination with recombinant vaccinia virus vaccines expressing glycoprotein genes of pseudorabies virus in the presence of maternal immunity. Vet Microbiol 58:93-103.

Brockmeier SL, Lager BCM, Tartaglia J, Riviere M, Paoletti E, Mengeling WL. 1993. Vaccination of pigs against pseudorabies with highly attenuated vaccinia (NYVAC) recombinant viruses. Vet Microbiol 38:41-58.

Brovra IH, Chakraverty P, Harris PA, Alexamder DJ. 1995. Disease outbreaks in pigs in Great Britain due to an influenza A virus of H1N2 serotype. Vet Rec 136:328-9.

Brown IH, Done SH, Spencer YI, Cooley WA, Harris PA, Alexander DJ. 1993. Pathogenicity of a swine influenza HlNl virus antigenically distinguishable from classical and European strains. Vet Rec 132:598-602.

BullerRML, Palumbo GJ. 1991. Poxvirus pathogenesis. Microbiol Rev 55:80-122.

Buller RML, Palumbo GJ. 1992. Safety and attenuation of vaccinia virus. Binns MM, Smith GL, eds. Recombinant Poxviruses. Boca Raton: CRC Press, p 235-67.

Buller RM, Smith GL, Moss B, Cremer K, Notkins AL. 1985. Infectious vaccinia virus Tk- recombinants that express foreign genes are less virulent than wild-type virus in mice. Lemer RA, Chanock RM, Brown F, eds. Vaccines 85. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, p 163-7. 37

Bullough PA, Hughson FM, Skehel JJ, Wiley DC. 1994. Structure of influenza haemagglutinin at the pH of membrane fijsion. Nature 371:37-43.

Cadoz M, Strady A, Meignier B, Taylor J, Tartaglia J, Paoletti E, Plotkin S. 1992. Immunisation with canarypox virus expressing rabies glycoprotein. Lancet 339:1429-32.

Carroll MW, Moss B. 1997. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus; propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238:198-211.

Carroll MW, Overwijk WW, Chamberlain RS, Rosenberg SA, Moss B, Restifo NP. 1997. Highly attenuated modified vaccinia virus Ankara (MVA) as an effective recombinant vector: a murine tumor model. Vaccine 15(4):387-94.

Cartwright FF, Biddiss MD. 1972. Disease and history. New York, NY: Dorset Press.

Castrucci MR, Donatelli I, Sidoli L, Barigazzi G, Kawaoka Y, Webster RG. 1993. Genetic reassortment between avian and human influenza A viruses in Italian pigs. Virology 193:503-6.

Chambers TM, Hinshaw VS, Kawaoka Y, Easterday BC, Webster RG. 1991. Influenza viral infection of swine in the United States 1988-89. Arch Virol 116:261-5.

Chang H-W, Uribe LH, Jacobs BL. 1995. Rescue of vaccinia virus lacking the E3L gene by mutants of E3L. J Virol 69:6605-8.

Chang H-W, Watson JC, Jacobs BL. 1992. The E3L gene of vaccinia virus encodes an inhibitor of the interferon-induced, double-stranded RNA-dependent protein kinase. Proc Natl Acad Sci U S A 89:4825-9.

Chelyapov NV, Antonova TP, Yanova NN, Chemos VI. 1988. Antigenic properties of vaccinia virus and of the virus recombinant strains expressing heterologous genes. Acta Virol (Praha) 32:409-16.

Claas ECJ, Kawaoka Y, de Jong JC, Masurel N, Webster RG. 1994. Infection of children with avian-human reassortant influenza virus from pigs in Europe. Virology 204:453-7.

Claas ECJ, Osterhaus ADME, van Beek R, De Jong JC, Rimmelzwaan GF, Senne DA, Krauss S, Shortridge KF, Webster RG. 1998a. Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus. Lancet 351:472-7.

Claas ECJ, de Jong JC, van Beek R, Rimmelzwaan GF, Osterhaus ADME. 1998b. Human influenza virus A/Hong Kong/156/97 (HlNl) infection. Vaccine 16(9/10):977-8. 38

Cooney EL, McElrath MJ, Corey L, Hu SL, et al. 1993. Enhanced immunity to HIV envelope elicited by a combined vaccine regimen consisting of priming with a vaccinia recombinant expressing HTV envelope and boosting with gpl60 protein. Proc Natl Acad Sci USA 90(5):1882-6.

Cooney EL, Collier AC, Greenberg PD, Coombs RW, Zarling J, Arditti DE, Hoffman MC, Hu S-L, Corey L. 1991. Safety of and immunological response to a recombinant vaccinia virus vaccine expressing HTV envelope glycoprotein. Lancet 337(8741);567-72.

Couch RB, Kasel JA. 1983. Immunity to influenza in man. Annu Rev Microbiol 37:529.

Couch RB, Keitel WA, Cate TR, Quarles JA, Taber LA, Glezen WP. 1996. Prevention of influenza virus infections by current inactivated influenza virus vaccines. Brown LE, Hampson AW, Webster RG, eds. Options for the control of influenza HI. Amsterdam: Elsevier, p 97-106.

Coupar BEH, Andrew ME, Boyle DB. 1988. A general method for the construction of recombinant vaccinia viruses expressing multiple foreign genes. Gene 68:1-10.

Cremer K, Mackett M, Wohlenberg C, Notkins AL, Moss B. 1985. Vaccinia virus recombinants expressing herpes simplex virus type 1 glycoprotein D prevents latent herpes in mice. Science 228:737-40.

Davison AJ, Moss B. 1989a. Structure of vaccinia virus early promoters. J Mol Biol 210:749-69.

Davison AJ, Moss B. 1989b. Structure of vaccinia virus late promoters. J Mol Biol 210:771-84. de St. Groth F, Webster RG. 1966. Disquisitions on original antigenic sin. I. Evidence in man. J Exp Med 124:331-45.

Dea S., Bilodeau R, Sauvageau R, Montpetit C, Martineau GP. 1992. Antigenic variant of swine influenza virus causing proliferative and necrotizing pneumonia in pigs. J Vet Diagn Invest 4:380-92.

Desmettre P, Languet B, Chappuis G, Brochier B, Thomas I, Lecocq J-P, Kieny M-P, Blancou J, Aubert M, Artois M, et al. 1990. Use of vaccinia rabies recombinant for oral vaccination of wildlife. Vet Microbiol 23:227-36.

Done SH, Brovm IH. 1997. Swine influenza in the United Kingdom, past and present. Large Animal Practice Mar/Apr:20-8. 39

Drillien R, Spehner D. 1983. Physical mapping of vaccinia virus temperature-sensitive mutations. Virology 131:385-93.

Easterday BC, Hinshaw VS. 1992. Swine influenza. Leman AD, Straw BE, Mengeling WL, D'Allaire S, Taylor DJ, eds. Diseases of swine. 7th ed. Ames, lA: Iowa State University Press, p 349-57.

Etlinger HM, Altenburger W. 1991. Overcoming inhibition of antibody responses to a malaria recombinant vaccinia virus caused by prior exposure to wild type virus. Vaccine 9:470-2.

Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. 1988. Smallpox and its eradication. Geneva: World Health Organization.

Flexner C, Murphy BR, Rooney JF, Wohlenberg C, Yuferov V, Notkins AL, Moss B. 1988. Successful vaccination with a polyvalent live vector despite existing immunity to an expressed antigen. Nature 335:259-62.

Francis T. 1953. Influenza: the newe acquayantance. Ann Intern Med 39(2);203-12.

Galletti R, Beauverger P, Wild TF. 1995. Passively administered antibody suppresses the induction of measles virus antibodies by vaccinia-measles recombinant viruses. Vaccine 13 (2):197-201.

Gammelin MJ, Mandler J, Scholtissek C. 1989. Two subtypes of nucleoproteins (NP) of influenza A viruses. Virology 170:71-80.

Gammelin M, Altmuller A, Reinhardt U, Mandler J, Harley VR, Hudson PJ, et ai 1990. Phylogenetic analysis of nucleoproteins suggests that human influenza A viruses emerged from a 19th-century avian ancestor. Mol Biol Evol 7:194-200.

Geraci JR, St. Aubin DJ, Barker IK, Webster RG, Hinshaw VS, Bean WJ. 1982. Mass mortality of harbor seals: Pneumonia associated with influenza A virus. Science 215:1129-31.

Gerhard W, Yewdell J, Frankel ME. 1981. Antigenic structure of influenza virus haemagglutinin defined by hybridoma antibodies. Nature 290:713-7.

Gillard S, Spehner D, Drillien R. 1985. Mapping of a vaccinia virus host range sequence by insertion into the viral thymidine kinase gene. J Virol 53:316-8.

Goebel SJ, Johnson GP, Perkus ME, Davis SW, Winslow JP, Paoletti E. 1990. The complete DNA sequence of vaccinia virus. Virology 179:247-66. 40

Gorman OT, Bean WJ, Kawaoka Y, Donatelli I, Guo Y, Webster RG. 1991. Evolution of influenza A virus nucleoprotein genes: implications for the origins of HlNl human and classical swine viruses. J Virol 65(7);3704-14.

Hinshaw VS, Alexander DJ, Aymard M, et ai 1984. Antigenic comparisons of swine- influenza-like HlNl isolates from pigs, birds and humans: an intemational collaborative study. Bull World Health Organ 62:871-8.

Hinshaw VS, Bean Jr. WJ, Webster RG, Easterday BC. 1978. The prevalence of influenza viruses in swine and the antigenic and genetic relatedness of influenza viruses from man and swine. Virology 84:51-62.

Hinshaw VS, Bean WJ, Geraci JR, Fiorelli P, Early G, Webster RG. 1986. Characterization of two influenza A viruses from a pilot whale. J Virol 58:655-6.

Hirsch VN, Fuerst TR, Sutter G, et ai 1996. Patterns of viral replication correllate with outcome in simian immunodeficiency virus-infected macaques: effects prior to immunization with a trivalent SIV vaccine in modified vaccinia vims Ankara (MVA). J Virol 70:3741-52.

Hodge JW, McLaughlin JP, Kantor JA, Schlom J. 1997. Diversified prime and boost protocols using recombinant vaccinia virus and recombinant non-replicatring avian pox virus to enhance T-cell immunity and antitumor responses. Vaccine 15(6/7):759-68.

Horimoto T, Rivera E, Pearson J, Senne D, Krauss S, Kawaoka Y, et ai 1995. Origin and molecular changes associated with emergence of a highly pathogenic H5N2 influenzavirus in Mexico. Virology 213:223-30.

Ito T, Nelson J, Couceiro SS, Kelm S, Baum LG, Krauss S, Castrucci MR, Donatelli I, Kida H, Paulson JC, et al. 1998. Molecular basis for the generation in pigs of influenza A viruses with pandemic potential. J Virol 72(9):7367-73.

Jenner E. 1798. An inquiry into the causes and effects of the variolae vaccinae. London; Sampson Low (Reprinted 1966. London: Dawsons of Pall Mall).

Johnson MP, Meitin CA Bender BS, Small Jr. PA. 1993. Recombinant vaccinia immunization in the presence of passively administered antibody. Vaccine ll(6):665-9.

Jones L, Ristow S, Yilma T, Moss B. 1986. Accidental human vaccination with vaccinia virus expressing nucleoprotein gene. Nature 319(6054):543.

Kaplan MM, Webster RG. 1977. The epidemiology of influenza. Sci Am 237:88-106. 41

Katz SL, Broome CV. 1991. Vaccinia (smallpox) vaccine. Recommendations of the immunization practices advisory committee (ACIP). Morbid. Mortal. Week. Rep. 40:1-10.

Kida H, Ito T, Yasuda J, Shimizu Y, Itakura C, Shortridge ECF, Kawaoka Y, Webster RG. 1994. Potential for transmission of avian influenza viruses to pigs. J Gen Virol 75:2183-8.

Kit S, Kit M, Pirtle EC. 1985. Attenuated properties of thymidine kinase-negative deletion mutant of pseudorabies virus. Am J Vet Res 46(6): 1359-67.

Kit S, Sheppard M, Ichimura H, Kit M. 1987. Second-generation pseudorabies vims vaccine with deletions in thymidine kinase and glycoprotein genes. Am J Vet Res 48(5):780-93.

Klenerman P, Zinkemagel RM. 1998. Original antigenic sin impairs cytotoxic T lymphocj^e responses to vimses bearing variant epitopes. Nature 394:482-5.

Konishi E, Pincus S, Paoletti E, Laegreid WW, Shope RE, Mason PW. 1992. A highly attenuated host range-restricted vaccinia vims strain, NYVAC, encoding the prM, E, and NS1 genes of Japanese encephalitis vims prevents JEV viremia in swine. Virology 190:454-8.

Kundig TM, Kalberer CP, Hengartaer H, Zinkemagel RM. 1993. Vaccination with two different vaccinia recombinant vimses: long-term inhibition of secondary vaccination. Vaccine 11(11):1154-8.

Kundin WD. 1970. Hong Kong A2 influenza vims infection among swine during a human epidemic in taiwan. Nature 228:857.

Lamb RA, Kmg RM. 1996. Orthomyxoviridae: The vimses and their replication. Fields BN, Knipe DM, Howley PM and others, eds. Fields Virology. 3rd ed. Philadelphia; Lippincott-Raven Publishers, p 1353-95.

Liang S, Mozdzanowska K, Palladino G, Gerhard W. 1994. Heterosubtypic immunity to influenza type A vims in mice. Effector mechanisms and their longevity. J Immunol 152:1653-61.

Luoh SM, McGregor MW, Hinshaw VS. 1992. Hemagglutinin mutations related to antigenic variation in HI swine influenza vimses. J Virol 66:1066-73.

Mayr A, Hochstein-Mintzel V, Stickl H. 1975. Abstammung, eigenschaften und verwendung des attenuierten vaccinia-stammes MVA. Infection 3:6-14. 42

Mayr A, Stickl H, Muller HK, Danner BC, Singer H. 1978. Der pockenimpfstamm MVA; Marker, genetische stniktur, erfahrungen mit der parenteralen schutzimpfling und verhalten im abwehrgeschwachten organismus. Zentralbl. Bakt. Hyg. I.(Abt. Orig. B) 167:375-90.

McMichael AJ, Gotch FM, Noble GR, Beare PAS. 1983. Cytotoxic T-cell immunity to influenza. N Engl J Med 309:13.

McMichael AJ. 1998. The original sin of killer T cells. Nature 394:421-2.

Meitin CA, Bender BS, Small Jr. PA. 1991. Influenza immunization: intranasal live vaccinia recombinant contrasted with parenteral inactivated vaccine. Vaccine 9:751-6.

Meitin CA, Bender BS, Small Jr. PA. 1994. Enteric immunization of mice against influenza with recombinant vaccinia. Proc Natl Acad Sci U S A 91:11187-91.

Meyer H, Sutter G, Mayr A. 1991. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J Gen Virol 72:1031-8.

Montefiori DC, Graham BS, Kliks S, Wright PF, et al. 1992. Serum antibodies to HTV-l in recombinant vaccinia virus recipients boosted with purified recombinant gpl60. J Clin Immunol 12(6):429-39.

Moss B. 1990. Regulation of vaccinia virus transcription. Aimu Rev Biochem 59:661-88.

Moss B. 1991. Vaccinia virus: a tool for research and vaccine development. Science 252:1662-7.

Murphy BR, Webster RG. 1996. Orthomyxoviruses. Fields BN, Knipe DM, Howley PM and others, eds. Fields Virology. 3rd ed. Philadelphia: Lippincott-Raven Publishers, p 1397-445.

Murphy FA. 1996. The Family Orthomyxoviridae. Fields BN, Knipe DM, Howley PM and others, eds. Fields Virology. 3rd ed. Philadelphia: Lippincott-Raven Publishers, p 36.

Nakamura RM, Easterday BC, Pawlisch R, Walker GL. 1972. Swine influenza: epizootiological and serological studies. Bull World Health Organ 47(4):481-7.

Oguiura N, Spehner D, Drillien R. 1993. Detection of a protein encoded by the vaccinia virus C7L open reading frame and study of its effect on virus multiplication in different cell lines. J Gen Virol 74:1409-13. 43

Ohishi K, Maruyama T, Shida H, Nishimaki J, Miki K, Sagata N, Dcawa Y, Sugimoto M. 1988. Immunogenicity of a recombinant vaccinia virus expressing envelope a glycoprotein of bovine leukaemia virus. Vaccine 6:428-32.

Olsen CW, McGregor MW, Cooley AJ, Schantz B, Hotze B, Elinshaw VS. 1993. Antigenic and genetic analysis of a recently isolated HlNl swine influenza virus. Am J Vet Res 54(10):1630-6.

Panicali D, Davis SW, Weinberg RL, Paoletti E. 1983. Construction of live vaccines by using genetically engineered poxviruses: biological activity of recombinant vaccinia virus expressing influenza virus hemagglutinin. Proc Natl Acad Sci U S A 80:5364-8.

Paoletti E, Lipinskas BR, Samsonoff C, Mercer S, Panicali D. 1984. Construction of live vaccines using geneticallly engineered poxviruses: biological activity of vaccinia virus recombinants expressing the hepatitis B virus surface antigen and the herpes simplex virus glycoprotein D. Proc Natl Acad Sci U S A 81:193-7.

Perkus ME, Goebel SJ, Davis SW, Johnson GP, Limbach K, Norton EE, Paoletti E. 1990. Vaccinia virus host range genes. Virology 179:276-86.

Perkus ME, Piccini A, Lipinskas BR, Paoletti E. 1985. Recombinant vaccinia virus: immunization against multiple pathogens. Science 229:981-4.

Pirtle EC, KQll HR, Swanson MR, Van Deusen RA. 1976. Hemagglutination-inhibiting antibodies against swine influenza and Hong Hong influenza viruses in swine sera in the USA. Bull World Health Organ 53:7-11.

Rekik MR, Arora DJS, Dea S. 1994. Genetic variation in swine influenza virus A isolate associated with proliferative and necrotizing pneumonia in pigs. J Clin Microbiol 32(2);515-8.

Renegar BCB, Small Jr. PA. 1991. Immunoglobulin A mediation of murine nasal anti- influenza virus immunity. J Virol 65(4):2146-8.

Rodriguez JF, Esteban M. 1989. Plaque size phenotype as a selectable marker to generate vaccinia virus recombinants. J Virol 63(2):997-1001.

Rohm C, Zhou N, Suss J, Mackenzie J, Webster RG. 1996. Characterization of a novel influenza hemagglutinin, HI 5: criteria for determination of influenza A subtypes. Virology 217(2):508-16.

Rooney JF, Wohlenberg C, Cremer KJ, Moss B, Notkins AL. 1988. Immunization with a vaccinia virus recombinant expressing herpes simplex virus type 1 glycoprotein D:long-term protection and effect of revaccination. J Virol (1530-1534). 44

Scheiflinger F, Falkner FG, Doraer F. 1996. Evaluation of the thymidine kinase (tk) locus as an insertion site in the highly attenuated vaccinia MVA strain. Arch Virol 141(663-669).

Schild GG, Newman RW, Webster RG, Major D, HQnshaw VS. 1980. Antigenic analysis of influenza A virus surface antigens: considerations for the nomenclature of influenza virus. Arch Virol 63:171-84.

Scholtissek C, Burger H, Bachmann PA, Hannoun C. 1983. Genetic relatedness of hemagglutinins of the HI subtype of influenza A viruses isolated from swine and birds. Virology 129:521-3.

Schultz U, Fitch WM, Ludwig S, Mandler J, Scholtissek C. 1991. Evolution of pig influenza viruses. Virology 183:61-73.

Senne DA, Panigrahy B, Kawaoka Y, Pearson JE, Suss J, Lipkind M, Kida H, Webster RG. 1996. Survey of the hemagglutinin (HA) cleavage site sequence of H5 and H7 avian influenza viruses: amino acid sequence at the HA cleavage site as a marker of pathogenicity potential. Avian Dis 40:425-37.

Shope RE. 1931. Swine influenza. lH. Filtration experiments and etiology. J Exp Med 54:373-80.

Shortridge BCF, Webster RG, Butterfield WK, Campbell CH. 1977. Persistence of Hong Kong influenza virus variants in pigs. Science 196:1454-5.

Small PAJr., Smith GL, Moss B. 1985. Intranasal vaccination with a recombinant vaccinia virus containing influenza hemagglutinin prevents both influenza virus pneumonia and nasal infection: intradermal vaccination prevents only viral pneumonia. Lemer RA, Chanock RM, Brown F, eds. Vaccines 88. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, p 175-6.

Smith GL, Bennink JR, Yewdell JW, Small PAJr., Murphy BR, Moss B. 1986. Vaccinia virus recombinants expressing influenza virus genes. Options for the control of influenza. Alan R. Liss, Inc.

Smith GL, Murphy BR, Moss B. 1983. Construction and characterization of an infectious vaccinia virus recombinant that expresses the influenza hemagglutinin gene and induces resistance to influenza virus infection in hamsters. Proc Natl Acad Sci USA 80:7155-9.

Suarez DL, Perdue ML, Cox N, Rowe T, Bender C, Huang J, Swayne DE. 1998. Comparisons of highly virulent H5N1 Influenza A viruses isolated from humans and chickens from Hong Kong. J Virol 72(8):6678-88. 45

Sutter G, Moss B. 1992. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc Natl Acad Sci USA 89:10847-51.

Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. 1994. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 12(11):1032-40.

Tartaglia J, Jarrett O, Neil JC, Desmettre P, Paoletti E. 1993. Protection of cats against feline leukemia virus by vaccination with a canarypox virus recombinant. J Virol 67:2370-5.

Tartaglia J, Perkus ME, Taylor J, Norton EK, Audonnet J-C, Cox WI, Davis SW, Van der Hoeven J, Meignier B, Riviere M, et al. 1992. NYVAC: a highly attenuated strain of vaccinia virus. Virology 188:217-32.

Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. 1997. Initial genetic characterization of the 1918 "Spanish" influenza virus. Science 275:1793-6.

Taylor CED. 1993. Did vaccinia virus come from a horse? Equine Vet J 25(1):8-10.

Taylor G, Stott EJ, Wertz G, Ball A. 1991. Comparison of the virulence of wild-type thymidine kinase (tk)-deficient and tk+ phenotypes of vaccinia virus recombinants after intranasal inoculation of mice. J Gen Virol 72:125-30.

Taylor J, Christensen L, Gettig R, Goebel J, Bouquet JF, Mickle TR, Paoletti E. 1996. Efficacy of a recombinant fowl pox-based Newcastle disease virus vaccine candidate against velogenic and respiratory challenge. Avian Dis 40(1): 173-80.

Taylor J, Weinberg R, Tartaglia J, Richardson C, Alkhatib G, Breidis D, et al. 1992. Nonreplicating viral vectors as potential vaccines: recombinant canarypox virus expressing measles virus fusion (F) and hemagglutinin (HA) glycoproteins. Virology 187:321-8.

Townsend ARM, Skehel JJ. 1984. The influenza A virus nucleoprotein gene controls the induction of both subtype specific and cross-reactive cytotoxic T cells. J Exp Med 160:552-63.

Traktman P. 1990. Poxviruses: an emerging portrait of biological strategy. Cell 62:621-6.

USD A. 1998. Veterinary Biological Products. Ames, LA: USD A/APHIS Center for Veterinary Biologies. 46 van der Leek ML, Feller JA, Sorensen G, Isaacson W, Adams CL, Borde DJ, PfeifFer N, Tran T, Meyer RW, Gibbs EP. 1994. Evaluation of swinepox virus as a vaccine vector in pigs using an Aujeszky's disease (pseudcrabies) virus gene insert coding for glycoproteins gp50 and gp63. Vet Rec 134(l):13-8. van Oirschot JT, Kaashoek MJ, Rijsewijk FA, Stegeman JA. 1996. The use of marker vaccines in eradication of herpesviruses. J Biotechnol 44:75-81.

Webster RG, Kawaoka Y, Taylor J, Weinberg R, Paoletti E. 1991. Efficacy of nucleoprotein and haemagglutinin antigens expressed in fowlpox virus as vaccine for influenza in chickens. Vaccine 9:303-8.

Webster RG, Taylor J, Pearson J, Rivera E, Paoletti E. 1996. Immunity to Mexican H5N2 avian influenza viruses induced by a fowl pox-H5 recombinant. Avian Dis 40(2):461-5.

Webster RG. 1998. Influenza: An emerging disease. Emerging Infectious Diseases 4(3):436-41.

Webster RG, Sharp GB, Claas ECJ. 1995. Interspecies transmission of influenza viruses. American J Respiratory and Critical Care Medicine 152;S25-S30.

Wentworth DE, Thompson BL, Xu X, Regnery HL, Cooley AJ, McGregor MW, Cox N, Hinshaw VS. 1994. An influenza A (HlNl) virus, closely related to swine influenza virus, responsible for a fatal case of human influenza. J Virol 68(4):2051-8.

WHO. 1980. A revision of the system of nomenclature for influenza viruses: a WHO Memorandum. Bull World Health Organ 58(4):585-91.

Woods GT. 1975. The prevalence of hemagglutination-inhibition (HI) antibodies to influenza vimses A/swine/Ill/63 and A/swine/Taiwan/7310/70 (H3N2) in Illinois swine herds, 1971-1973. Res Commun Chem Pathol Pharmacol 10:573-6.

Wraith CD, Vessey AE, Askonas BA. 1987. Purified influenza virus nucleoprotein protects mice from lethal infection. J Gen Virol 68:433-40.

Wyatt LS, Shors ST, Murphy BR, Moss B. 1996. Development of a replication-deficient recombinant vaccinia virus vaccine effective against parainfluenza virus 3 infection in an animal model. Vaccine 14(15): 1451-8.

Yetter RA, Lehrer S, Ramphal R, Small Jr. PA. 1980. Outcome of influenza infection:effect of site of initial infection and heterotypic immunity. Infect Immun 29(2):654-62. 47

Yewdell JW, Bennink JR, Smith GL, Moss B. 1985. Influenza A virus nucieoprotein is a major target antigen for cross-reactive anti-influenza A virus cytotoxic T lymphocytes. Proc Natl Acad Sci USA 82:1785-9.

Yuen KY, Chan PKS, Peiris M, Tsang DNC, Que TL, Shortridge KF, Cheung PT, To WK, Ho ETF, Sung R, et ai 1998. Clinical features and rapid viral diagnosis of human disease associated with avian influenza A H5N1 virus. Lancet 351:467-71. 48

I. EVALUATION OF IMMUNOGENICITY, SAFETY AND HOST RESTRICTION OF THE HIGHLY ATTENUATED MVA STRAIN OF VACCINIA VIRUS EXPRESSING HUMAN INFLUENZA VIRUS HEMAGGLUTININ AND NUCLEOPROTEIN GENES

A paper to be submitted to Vaccine

Patricia L. Foley, Steven K. Hanson, Randall L. Levings, and Prem S. Paul

ABSTRACT

The highly attenuated modified vaccinia Ankara (MVA) strain of vaccinia virus was used as a vector for the delivery of influenza virus hemagglutinin (HA) and nucleoprotein (NP) genes from the HlNl human isolate, A/PR/8/34 (PR/8). The level of antigenic relatedness and serologic cross-reactivity between the HA and NP genes of PR/8, the MVA recombinant

(MVA-HA-NP), and various swine and turkey influenza virus isolates was examined. It was determined that, serologically, the HlNl swine and turkey isolates are closely related compared to the human PR/8 HlNl strain. In addition, twenty pigs, or five groups of four pigs each, were challenged intranasally with 10* EID50 of Sw/IN/1726/88 propagated in embryonated chicken eggs. Two groups had been previously vaccinated twice with the M\''A recombinant, one group intramuscularly and the other intranasally. Another two groups were vaccinated similarly with the parental MVA strain. The fifl:h group remained unvaccinated.

Pigs vaccinated with the MVA recombinant had significantly less duration and lower titer of viral shedding, compared to those vaccinated with the parental strain and nonvaccinated 49

challenged controls. This protection was presumably due to cell-mediated immune response

generated against NP, since humoral response to both HA and NP was undetectable up to the

day of challenge. All pigs had comparable titers to swine influenza virus HA at 7 days post

challenge, whereas those vaccinated wnth the recombinant virus had significantly greater

serologic response to NP when compared to the controls. Also, it was determined that

MVA-HA-NP was avirulent for mice when given by intracranial, intraperitoneal, or subcutaneous route of inoculation. The host restriction of MVA strain replication, previously reported for numerous mammalian species, was confirmed in vitro through immunostaining of porcine cell culture.

INTRODUCTION

Swine influenza (SI) is a common disease of pigs caused by type A influenza viruses, mostly of the HlNl antigenic subtype It is characterized by pyrexia, anorexia, dyspnea, coughing, sneezing, depression, huddling, and pneumonia. In acutely affected herds, it exhibits 100% morbidity, although mortality is low. SI is likely to continue on in a herd that is not depopulated, causing recurring outbreaks of respiratory disease and reproductive problems. With an estimated one-third of pigs in the U. S. infected with SI economic impact on the swine industry is considerable. In addition, there are ongoing public health concerns regarding the transmission of SI to humans, which can on occasion result in death

In the U. S., SI virus (SIV) strains have exhibited little antigenic variation over time, with 50 one dominant antigenic type responsible for most disease outbreaks. This phenomenon has

been attributed to the lack of selection pressure provided by the relatively short lifespan of swine and the continual presence of a susceptible population, i.e. young pigs without antibodies It has previously been demonstrated that monoclonal antibodies generated against the human A/PR/8/34 (PR/8) strain are cross-reactive with antigenic sites on

A/Swine/Iowa/31 (IA/31), a classic example of the dominant antigenic type of SIV in pigs

It has also been shown that the hemagglutinin (HA) of A/Sw/In/1726/88 (IN/88) has three antigenic sites that correspond to the three sites on the HA of PR/8, located in the loop and distal tip of the protein These findings suggest that some degree of cross reactivity will occur between antibodies generated against the PR/8 strain and various sv^e influenza isolates. If so, the immune response induced by a recombinant virus expressing important

PR/8 immunogens may protect pigs against SFV. In this study, we evaluated the ability of a recombinant vaccinia vector containing influenza virus HA and nucleoprotein (NP) genes fi-om the PR/8 strain to provide immunity to SW.

This recombinant vaccinia vector and the modified vaccinia virus Ankara (MVA) strain ft-om which it is derived have been well characterized 34.35 ^^s been used without significant side-effects in over 120,000 people for immunization against smallpox It contains six major deletions that prevent virus assembly in most all mammalian cells tested but leave gene expression, both early and late, intact. This host restriction, although demonstrated in many mammalian cell lines has only been examined in one cell line of swine. In this study, we determined its restriction to extend to at least three cell types. Moreover, we evaluated its ability to pass the mouse safety test required for 51 licensure of live virus vaccines, as described in the U. S. Code of Federal Regulations,

Chapter 9, part 113.33(a)".

The recombinant MVA vector contains the HA and NP genes from PR/8 in opposite orientation and under the control of two optimized synthetic promoters, and has been shown to provide protection to mice against a lethal challenge from PR/8 Previous studies have demonstrated that HA can induce virus-neutralizing antibody however, NP induces only non-neutralizing antibody NP, however, is the major target antigen recognized by type- specific cross-reactive cytotoxic T-lymphocytes (CTL) whereas HA only stimulates subtype-specific CTL It is known that antibody is important for protection against infection and that CTL response accounts for recovery from infection^'. Since most of the protective humoral response following exposure to influenza virus is generated against the five hypervariable antigenic sites of HA, antibodies to this protein provide immunity to influenza viruses with a sufflciently similar HA We examined antisera to MVA-HA-NP,

PR/8, and various isolates of SIV to determine in vitro their ability to react with the HA of

PR/8, IN/88, and several avian influenza virus (ATV) isolates from turkeys. We found that, serologically, the swine and turkey isolates were more closely related. In addition, sera from swine vaccinated with MVA-HA-NP then challenged with IN/88 were evaluated through 7 days postchallenge for their reactivity in a hemagglutination inhibition (HAI) assay, in a competitive ELISA (cELISA) with influenza type A antibody to NP, and in an agar gel immunodifiusion (AGED) assay with type A NP and matrix protein antibody. The humoral responses to HA and NP were compared. 52

MATERIALS AND METHODS

Antisera

One porcine-origin hyperimmune serum generated against a 1960's SIV field isolate, provided by the Diagnostic Virology Laboratory (DVL), National Veterinary Services

Laboratories (NVSL), and another ferret-origin anti-PR/8 hyperimmune serum, provided by the Influenza Division, Centers for Disease Control and Prevention (CDC), were compared in an HAI assay for their ability to neutralize the respective homologous virus, as well as IA/31, a more recent SIV isolate, and several HlNl AIV isolates fi'om turkey farms in Minnesota

(MN/88), Missouri (MO/87), and South Dakota (SD/86). Serum samples were also collected fi-om 1-, 24-, and 40-day old pigs in a conventional swine herd and tested similarly.

Four caesarean-derived colostrum-deprived (CDCD) pigs were challenged with 1 ml per nostril of 10^ mean egg infectious dose (EID5o)/ml IN/88; their HAI titers were evaluated at 7 days post challenge against the homologous virus and the avian isolates. Two gnotobiotic pigs were inoculated IM with 10® fluorescent focal units (FFU) of MVA-HA-NP at 14 days of age and again at 37 days. Serum titers at 14, 35, and 51 days of age were evaluated against virus isolates PRy8, IN/88, MN/88, MO/87, and SD/86.

Vinises

Vaccinia virus strains MVA, MVA-HA-NP, and the nonattenuated, nonrestricted Western

Reserve (WR) recombinant, WR-HA-NP, were generously provided by B. Moss, Laboratory for Viral Diseases (LVD), National Institute of Allergy and Infectious Diseases (NIAID), 53

National Institutes of Health (NIH). MVA and MVA-HA-NP were grown and titrated by

fluorescent focal assay in chicken embryo fibroblast cells, using Minimum Essential Medium

(MEM, Gibco BRL) and 5% fetal bovine serum (FBS), and expressed as FFU. WR-HA-NP

was propagated on HeLa cells and titrated by plaque assay on BS-C-1 cells, with titers

expressed as plaque-forming units (PFU). The influenza virus PR/8, provided by the

Influenza Division, CDC, was propagated in the allantoic sac of 10-day-old embryonated

chicken eggs at 37° C for 72 hours. IA/31 and a 1960's SIV field isolate were acquired fi'om

the DVL, NVSL, for use in HAI assays. The challenge virus, IN/88, was provided by

Virginia Hinshaw, University of Wisconsin. The virus was grown in the allantoic cavity of

10-day-old embryonated chicken eggs at 37°C for 72 hours, and its EID50 and HA unit titers

were lO^ Vml and 128, respectively.

Mouse safety tests

Using nine groups of eight mice each, the MVA recombinant (10^-^ FFU/ml), the MVA

parental strain (10^-^ FFU/ml), or the WR recombinant (10®-^ PFU/ml) was inoculated into

mice using either the intracranial (0.03 ml/mouse), intraperitoneal (0.5 ml/mouse), or subcutaneous (0.5 ml/mouse) route of inoculation. Six mice served as uninoculated controls.

Mice were observed for seven days.

Immunoassay

Cells were maintained in a humidified, 5% CO, atmosphere at 37° until the development of a complete cell monolayer. ST cells (a swine testicular cell line) and PK-15 cells (a 54

porcine kidney cell line) were grown in Eagle's MEM medium supplemented with 1% L- glutamine, 1% sodium pyruvate, and 5% FBS. SKP cells (swine kidney primary cells) were

propagated in Eagle's MEM medium supplemented with 1% L-glutamine, 1% sodium pyruvate, and 10% FBS. Following viral inoculation at dilutions of 10'~ and 10'^, at least 10^ and 10^ respectively of each virus, the cells were maintained in similar media but containing only 2.5% FBS. At 24 or 48 hours after virus inoculation, the cells were incubated with an anti-PR/8 HA mouse monoclonal antibody (courtesy of B. Moss, LVD, NIAID, NIH) and anti-mouse IgG antibodies conjugated to peroxidase, then stained with o-dianisidine.

Vaccination/challenge

At 21 days of age, four groups of four caesarean-derived colostrum-deprived (CDCD) pigs were vaccinated with 10* FFU of the MVA parent or the MVA recombinant either intramuscularly (IM) or intranasally (IN). This was repeated at 35 days of age. At 49 days of age, these sixteen pigs and four unvaccinated controls were challenged intranasally, 1 ml per nares, with 10®-^ EID50 of strain IN/88 harvested from allantoic fluid of embryonated eggs. The eggs had been inoculated at 10 days with virus isolated from infected pigs. Nasal swabs were collected at 2, 3, 4, 5, and 7 days post-challenge and titrated in embryonated chicken eggs to determine levels of virus shedding.

Virus isolation and titration

Nasal swab samples were placed in tubes containing 2 mis Eagle's MEM supplemented with 75 U/ml penicillin G potassium, 225 U/ml streptomycin sulfate, 0.1% gentocin, and 55

1.5% amphotericin B. The tubes were frozen, then thawed, vortex-mixed, and centrifuged at

400xg for 10 minutes. The allantoic sacs of ten-day-old embryonated eggs were inoculated with 0.1 ml of the resulting supernatant and incubated at 37° for 72 hr. The harvested allantoic fluids were tested for HA activity. Eggs were inoculated in duplicate to determine the presence of virus. For titration, four eggs were inoculated v«th each dilution, up to 10"^, and the results tabulated using the method of Spearman-Karber

Serological assays

HA and HAI tests were performed in 96-well microtiter plates, using 0.5% chicken erythrocytes. For the SIV and PR/8 HAI assay, sera were pretreated using 10% kaolin and

5% washed chicken erythrocytes, then evaluated at 1:10 or greater dilutions against live virus antigen. For the ATV HAI assay, virus inactivated with 0.1% beta-propiolactone served as antigen; serial dilutions of antisera began at 1:8. For both assays, four HA units of each virus were used to determine serum HAI titers. The AGID assay, used to detect circulating antibodies to two type A influenza group-specific antigens, NP and matrix protein (M), was conducted as described by Beard'^ using A/TY/MN/3689-I551/81 H5N2 virus to produce standardized antigen and antiserum. The type A mfluenza cELISA procedure used was similar to that described by Katz et al for vesicular stomatitis virus and made use of type A

NP recombinant baculovirus antigen cloned from A/Arm Arbor/6/60 type A-specific anti-

NP monoclonal antibody H16 peroxidase-labeled conjugate, ABTS substrate, and standardized specific reference antisera. 56

Statistical analysis

Data was evaluated using an analysis of variance and then the Least Significant Difference method to compare means.

RESULTS

Cross reactivity between influenza vims isolates

A low level of HAI cross-reaction between PR/8 antisera and SIV and AIV isolates was found using hyperimmune sera (Table 1). The PR/8 hyperimmune sera had the highest titers to PR/8. The DVL antisera, however, did not have a detectable titer to PR/8, but did cross react with the AIV isolates to differing degree. The conventional herd samples demonstrated titers to SIV and AIV that, in one instance, cross-reacted with PR/8.

Sera from CDCD pigs challenged with SIV demonstrated variability of response to AIV antigen at 7 days postchallenge (Table 2). The SD/86 and MN/88 isolates elicited stronger reaction than MO/87.

Testing of the gnotobiotic sera (Table 3) showed that vaccination with MVA-HA-NP resulted in high HAI titers to PR/8, up to 1:1280, but response to IN/88 was not detectable.

Mouse safety

None of the mice inoculated with the MVA recombinant or the MVA parental strain intraperitoneaUy, subcutaneously, or intracranially, or the WR recombinant intraperitoneally or subcutaneously, were clinically affected. However, the eight mice that received 57 intracranial inoculation of the WR recombinant were clinically affected. In this group, two died and six exhibited weakness and rough coats during the 7-day observation period.

Susceptibility of cells to vaccinia virus replication

All three cell types tested demonstrated host restriction, in that none had iVTVA-HA-NP- or MVA-induced lytic cytopathic effect typical of cells fiilly permissive to vaccinia virus replication (Figure 1). The ST and PK-15 lines were also somewhat restrictive to the WR-

HA-NP recombinant, in that the expected cell lysis did not occur in these cell lines.

However, the WR recombinant induced rapid cytopathic effect and cell lysis in the SKP primary cells. Single-cell foci of protein expression were evident for both the recombinant strains after immunostaining of infected ST and PK-15 cell monolayers. The stain was more pronounced after 48 hours of viral growth in ST and PK-15 cells than after 24 hours in SKPs.

Efficacy of the recombinant vaccine

Vaccination with the MVA recombinant did not prevent infection but did provide some protection against the amount and duration of virus shedding (Table 4). Use of MVA alone, given IM or EN, had an apparent protective effect compared to the nonvaccinated challenged controls, but the differences were not significant. Clinical signs, following challenge, were not pronounced nor sufiBciently different to distinguish between MVA-HA-NP vaccinates and other animals, except on Day 2 postchallenge. On day 2 the nonvaccinated controls displayed greater signs of mild to moderate dyspnea (3 out of 4), sneezing (2/4), nasal discharge (2/4), coughing (1/4), and elevated temperature (1/4) compared to the other four 58 groups after challenge with virulent influenza virus. Postmortem examination revealed that all animals except one receiving MVA-HA-NP intranasally had gross and histopathological lesions indicative of SIV infection.

Following each vaccination, MVA-HA-NP induced higher humoral response to the PR/8

HA, especially for the IM group (Table 5). However, there was no detectable serologic response to HA of IN/88 throughout the vaccination period. At 7 days postchallenge, HAI titers of the vaccinates to IN/88 were comparable to those of the nonvaccinated challenged controls. The MVA-HA-NP vaccinates demonstrated a stronger immune response on cELIS A for type A NP antibody, as well as on the AGED for NP and M antibody, when compared to other groups. This suggests that the challenge elicited a booster effect to the initial vaccinations.

Statistical analysis

Use of the recombinant vaccine by both EM and IN routes resulted in significantly fewer days of shedding following SIV challenge (p<0.05) when compared to all groups except that given the parent MVA strain by IM route (p>0.08) (Table 6). Titers of shed virus dropped significantly for the group vaccinated IN with the recombinant starting on day 4 postchallenge, and on day 5 for the group receiving the recombinant IM (p<0.05) (Tables 7,

8). With only five animals per treatment group, small but significant differences between groups may have been missed.

Seven days after SIV challenge, animals that had been vaccinated with MVA-HA-NP had significantly higher antibody titer to type A NP when compared to the other groups (Table 59

9). In contrast, HAI titers against the IN/88 strain of SIV were similar in vaccinated and nonvaccinated pigs.

DISCUSSION

The results of this study are consistent with previous reports regarding the role of NP- generated immune response during infection with influenza virus Expression of influenza virus type A-specific NP by the recombinant MVA-HA-NP did not prevent infection but assisted in more rapid recovery from infection. Antibody to PR8 HA did not protect against the heterologous SIV challenge. Our study also suggests that both the MVA parent and the recombinant MVA-HA-NP can be safely administered to pigs and mice at high doses by various routes. Even though both parent and recombinant virus are host-restricted in the three tj'pes of porcine cells tested, MVA-HA-NP apparently expresses sufficient protein to induce a protective effect in the pig, even to heterologous challenge. Presumably, this is due to NP's enhancement of cell-mediated immunity, since there was no detectable humoral response to HA prior to challenge nor an apparent booster response following challenge, as was seen in the AGID and cELISA tests for NP. If NP-specific CTLs did play a part in viral clearance, as has been demonstrated previously, the response was not rapid enough to prevent initial infection. Replacement of the PR/8 HA and NP genes with those of IN/88 should result in a superior vaccine, since previous studies indicate that HA stimulates the needed initial protection against homologous influenza challenge

It has been reported that mice vaccinated against influenza with a vaccinia virus-HA 60 recombinant given by systemic routes (IP, scarification) developed serum antibody that protected against lung but not nasal infection. In contrast, intranasal immunization provided both serum and nasal antibody and protection against infection at both sites This has been attributed to the effect of IgA nasal antibody generated by IN inoculation of virus. We investigated the difference between IM and IN inoculation and noted a more rapid clearance of virus shedding fi^om the nose in pigs given MVA-HA-NP by nasal route. This presumably was not due to HA-specific IgA, since antibody to the recombinant's PR/8-origin HA did not react with SIV HA at a detectable level on the HAI assay. It is possible that, although not detected in vitro, a low level of IgA antibody was present in vivo to offer some local protection.

It is believed that all mammalian influenza viruses are derived from an avian influenza virus but that presently aquatic birds and humans maintain two large and distinct reservoirs Swine can be infected by strains from either source Moreover, swine are believed to be the primary species generating reassortants and serving as a conduit from one reservoir to another for the influenza virus. Phylogenetically, the NP genes of PR/8 and classical SIV isolates are closely related But, the NP as well as HA genes of influenza viruses found recently in European pigs have been of avian origin These 'avianized' SIVs are believed to have arisen by reassortment in pigs of genes from avian and human influenza virus strains and have themselves, in turn, infected humans and turkeys In one study of

73 swine isolates from 11 states in the United States (U. S.), collected from 1976 to 1990, and II turkey isolates from 8 states, collected from 1980 to 1989, all genes analyzed from the strains infecting pigs were characteristic of SIV, suggesting that U. S. pigs are not all that 61 frequently involved in genetic exchange However, 73% of the turkey isolates contained genes of swine origin, implying that turkeys in the U. S. can and do serve as frequent recipients of gene transfer. Our findings are consistent with this hypothesis. Whereas antisera to classic SIV strains reacted well with both SIV and turkey isolates on HAI assay, there is no corresponding response to the human strain, PR/8. Conversely, antisera to PR/8 or its recombinant antigens did not react with the SIV or ATV antigens studied. This raises the possibiUty of developing a single vaccine able to protect both pigs and domestic birds against certain strains of influenza virus.

References

1. Altenburger W, Suter C-P, Altenburger J. Partial deletion of the human host range gene in the attenuated vaccinia virus MVA. Arch Virol 1989;105:15-27.

2. Andrew ME, Coupar BEH, Ada GL, Boyle DB. Cell-mediated immune response to influenza virus antigens expressed by vaccinia virus recombinants. Microb Pathos 1986; 1:443-52.

3. Andrew ME, Coupar BEH, Boyle DB, Ada GL. The roles of influenza virus haemagglutinin and nucleoprotein as protective antigens against influenza virus infection in mice. Scand J Immunol 1987;25;21-8.

4. Antoine G, Scheiflinger F, Holzer G, Langmann T, Falkner EG, Domer F. Characterization of the vaccinia MVA hemagglutinin gene locus and its evaluation as an insertion site for foreign genes. Gene 1996;177:43-6.

5. Ben-Yehuda A, Ehleiter D, Hu A, Weksler ME. Recombinant vaccinia virus expressing the PR/8 influenza hemagglutinin gene overcomes the impaired immune response and increased susceptibility of old mice to influenza infection. J Infect Dis 1993;168:352-7. 62

6. Bender BS, Rowe CA, Taylor SF, Wyatt LS, Moss B, Small Jr. PA. Oral immunization with a replication-deficient recombinant vaccinia virus protects mice against influenza. J Virol 1996 Sep;70(9):6418-24.

7. Bennink JR, Yewdell JW, Smith GL, Moller C, Moss B. Recombinant vaccinia virus primes and stimulates influenza haemagglutinin-specific cytotoxic T cells. Nature 1984 Oct;311(5986):578-9.

8. Carroll MW, Moss B. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus; propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 1997;238:198-211.

9. Chambers TM, Hinshaw VS, Kawaoka Y, Easterday BC, Webster RG. Influenza viral infection of swine in the United States 1988-89. Arch Virol 1991; 116:261-5.

10. Claas ECJ, Kawaoka Y, de Jong JC, Masurel N, Webster RG. Infection of children with avian-human reassortant influenza virus from pigs in Europe. Virology 1994;204:453-7.

11. Easterday BC, Hinshaw VS. Swine influenza. Leman AD, Straw BE, Mengeling WL, D'Allaire S, Taylor DJ, eds. Diseases of swine. 7th ed. Ames, lA: Iowa State University Press; 1992. p 349-57.

12. Finney DJ. Spearman-Karber and moving averages. Statistical method in biological assay. 3rd ed. London: Charles Griffin & Company Ltd; 1978. p 394-401.

13. Gerhard W, Yewdell J, Frankel ME. Antigenic structure of influenza virus haemagglutinin defined by hybridoma antibodies. Nature 198 l;290:7I3-7.

14. Hemphill J, ed. Title 9 - Animals and animal products. Chapter 1 (parts 1- 199). Code of federal regulations. Washington D. C. U. S. Government Printing Office; 1995.

15. Hinshaw VS, Bean Jr. WJ, Webster RG, Easterday BC. The prevalence of influenza viruses in swine and the antigenic and genetic relatedness of influenza viruses from man and swine. Virology 1978;84:51-62.

16. Karber G. Beitrag zur kollektiven behandlung pharmakologischer reihenversuche. Arch Exp Pathol Pharmakol 1931; 162:480-3.

17. Katz JB, Shafer AL, Eemisse KA. Construction and insect larval expression of recombinant vesicular stomatitis nucleocapsid protein and its use in competitive ELISA. J Virol Methods 1995;54:145-57. 63

18. Kida H, Ito T, Yasuda J, Shimizu Y, Itakura C, Shortridge BCF, Kawaoka Y, Webster RG. Potential for transmission of avian influenza viruses to pigs. J Gen Virol 1994;75:2183-8.

19. Ludwig S, Haustein A, Kaleta EF, Scholtissek C. Recent influenza A (HlNl) infections of pigs and turkeys in northern Europe. Virology 1994;202;281-6.

20. Luoh SM, McGregor MW, Hinshaw VS. Hemagglutinin mutations related to antigenic variation in HI swine influenza viruses. J Virol 1992;66:1066-73.

21. Mayr A, Hochstein-Mintzel V, Stickl H. Abstammung, eigenschaften und verwendung des attenuierten vaccinia-stammes MVA. Infection 1975;3:6-14.

22. Mayr A, Stickl H, Muller HEC, Danner EC, Singer H. Der pockenimpfstamm MVA: Marker, genetische struktur, erfahrungen mit der parenteralen schutzimpfung und verhalten im abwehrgeschwachten organismus. Zentralbl. Bakt. Hyg. I.(Abt. Orig. B) 1978;167:375-90.

23. Meitin CA, Bender BS, Small Jr. PA. Influenza immunization: intranasal live vaccinia recombinant contrasted with parenteral inactivated vaccine. Vaccine 1991 Oct; 9:751-6.

24. Meyer H, Sutter G, Mayr A. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J Gen Virol 1991;72:1031-8.

25. Pirtle EC, ffill HR, Swanson MR, Van Deusen RA. Hemagglutination-inhibiting antibodies against swine influenza and Hong Hong influenza viruses in swine sera in the USA. Bull World Health Organ 1976;53:7-11.

26. Rota PA, Black RA, De BK, Harmon MW, Kendal AP. Expression of influenza A and B virus nucleoprotein antigens in baculovirus. J Gen Virol 1990;71:1545-54.

27. Scheiflinger F, Falkner FG, Domer F. Evaluation of the thymidine kinase (tk) locus as an insertion site in the highly attenuated vaccinia MVA strain. Arch Virol 1996;141(663-669).

28. Schmidt NJ, Emmons RW. General principles of laboratory diagnostic methods for viral, rickettsial, and chlamydial infections. Schmidt NJ, Emmons RW, eds. Diagnostic procedures for viral, rickettsial, and chlamydial infections. 6th ed. Washington, D. C.: American Public Health Association; 1989. p 1-35.

29. Scholtissek C, Ludwig S, Fitch WM. Analysis of influenza A virus nucleoproteins for the assessment of molecular genetic mechanisms leading to new phylogenetic virus lineages. Arch Virol 1993;131:237-50. 64

30. Schultz U, Fitch WM, Ludwig S, Mandler J, Scholtissek C. Evolution of pig influenza viruses. Virology 1991;183:61-73.

31. Shu LL, Lin YP, Wright SM, Shortridge BCF, Webster RG. Evidence for interspecies transmission and reassortment of influenza A viruses in pigs in southern China. Virology 1994;202:825-33.

32. Small PAJr., Smith GL, Moss B. Intranasal vaccination with a recombinant vaccinia virus containing influenza hemagglutinin prevents both influenza virus pneumonia and nasal infection: intradermal vaccination prevents only viral pneumonia. Lemer RA, Chanock RM, Brown F, eds. Vaccines 88. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1985. p 175-6.

33. Smith GL, Murphy BR, Moss B. Construction and characterization of an infectious vaccinia virus recombinant that expresses the influenza hemagglutinin gene and induces resistance to influenza virus infection in hamsters. Proc Natl Acad Sci U S A 1983 Dec;80:7155-9.

34. Sutter G, Moss B. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc Natl Acad Sci USA 1992 Nov;89;10847-51.

35. Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 1994; 12(11): 1032-40.

36. Towiisend ARM, Skehel JJ. The influenza A virus nucleoprotein gene controls the induction of both subt>'pe specific and cross-reactive cytotoxic T cells. J Exp Med 1984;160:552-63.

37. Webster RG, Kawaoka Y, Taylor J, Weinberg R, Paoletti E. Efficacy of nucleoprotein and haemagglutinin antigens expressed in fowlpox virus as vaccine for influenza in chickens. Vaccine 1991;9:303-8.

38. Webster RG, Sharp GB, Claas ECJ. Interspecies transmission of influenza viruses. American J Respiratory and Critical Care Medicine 1995;152:S25-S30.

39. Wentworth DE, Thompson BL, Xu X, Regnery HL, Cooley AJ, McGregor MW, Cox N, Hinshaw VS. An influenza A (HlNl) virus, closely related to swine influenza virus, responsible for a fatal case of human influenza. J Virol 1994;68(4);2051-8.

40. Woods GT. The prevalence of hemagglutination-inhibition (HI) antibodies to influenza viruses A/swine/Ill/63 and A/swine/Taiwan/7310/70 (H3N2) in Illinois swine herds, 1971-1973. Res Commun Chem Pathol Pharmacol 1975;10:573-6. 65

41. Wraith CD, Vessey AE, Askonas BA. Purified influenza virus nucleoprotein protects mice from lethal infection. J Gen Virol 1987;68:433-40.

42. Wright SM, Kawaoka Y, Sharp GB, Senne DA, Webster RG. Interspecies transmission and reassortment of influenza A viruses in pigs and turkeys in the United States. Am J Epidemiol 1992;136:488-97.

43. Yewdell JW, Frank E, Gerhard W. Expression of influenza A virus internal antigens on the surface of infected P815 cells. J Immunol 1981;126:1814-9.

44. Yewdell JW, Bennink JR, Smith GL, Moss B. Influenza A virus nucleoprotein is a major target antigen for cross-reactive anti-influenza A virus cytotoxic T lymphocytes. Proc Natl Acad Sci U S .A 1985 Mar;82:1785-9. Table 1. Serologic cross reactivity of antisera against various influenza virus isolates.

Serum HAl tilers vs. 7 different HINI influenza virus isolatesB

Immunizing , IA/31 8750 PR/8 MN/88 SD/86 MO/87 virus ^ animal type DVL DVL pig hyperimmune S 1280 S 1280 S 1280 neg 1024 512 128

PR/8 ferret hyperimmune 10 10 20 ^ 1280 16 8 8 unknown 1 day old pig conventional 40 40 40 20 32 32 neg unknown 24 day old pigs'' conventional 13 13 10 neg 23 36 6

^HAI titers are expressed as the reciprocal of the last dilution of antisera inhibiting 4 HA units of virus, DVL = 1960's SIV isolate, I/W31 = 1931 SIV isolate, 8750 = 1990's SIV isolate; SD/86, MO/87, and MN/88 = 1980's turkey influenza virus isolates. ^These titers, which indicate maternal immunity, are geometric mean titers (GMT) for six 24-day-old conventional pigs. The 1-day-old pig also had passive immunity to these isolates.

neg=negative. For PR8 HAI assay, neg<10; for AIV HAl assay, neg<8. 67

Table 2. Reactivity of sera from pigs challenged with Sw/IN/88 against various i^Quenza virus isolates as demonstrated by HAI.

Days post HAI titers against influenza virus isolates* Hg No. chaHenge Sw/IN/88 Ty/SD/86 Ty/MO/87 Ty/MN/88

68 0 neg neg neg neg 71 0 neg neg neg neg 74 0 neg neg neg neg 125 0 neg neg neg neg

68 7 20 256 8 64 71 7 20 64 16 32 74 7 40 256 16 128 125 7 40 64 16 128

*Reciprocal of last dilution which inhibits 4 HA units of influenza virus. neg = negative. For STV HAI assay, neg <10; for ATV HAI assay, neg <8. Lower titers to homologous SIV are due to procedural differences in the 2 assays. 68

Table 3. Antibody response of pigs to recombinant virus expressing HA and MP at various times after vaccination.

Days post HAI titers against influenza virus isolates* Pig No. Age (days) vaccinatioii #1 PR/8 IN/88 Ty/MO/87 Ty/MN/88 Ty/SD/86

6 14 0 neg neg neg neg neg 7 14 0 neg neg neg neg neg 6 35 21 80 neg neg neg neg 7 35 21 160 neg neg neg neg 6 51 37 640 neg neg neg neg 7 51 37 1280 neg neg neg neg

* Titers are expressed as the reciprocal of the last dilution which inhibits 4 HA units of influenza virus. Two gnotobiotic pigs were vaccinated IM with MVA-HA-NP on 14 and 37 days of age.

neg = negative. For SIV" HAI assay, neg <10; for ATV HAI assay, neg <8. Figure 1 Immunostaining for influenza virus proteins in cell cultures infected with vaccinia virus expressing influenza virus proteins. Column A contains ST cells at 48 hours postinoculation. Column B contains PK-15 cells 48 hours postinoculation. Column C contains swine kidney cells 24 hours postinoculation. Row I, from left to right, consists of uninoculated cell controls. In row 2, each of the cultures was inoculated with the MVA strain of vaccinia virus. In row 3, the cultures were infected with the MVA/PR8 recombinant. Lastly, the cultures in row 4 were infected with the nonattenuated Western Reserve strain of vaccinia virus containing the PR/8 HA and NP genes. All monolayers were stained with anti-influenza virus HA mouse monoclonal antibody and anti-mouse IgG antibodies conjugated to peroxidase. Color developed, following staining with o-dianisidine, only in the bottom two rows, confirming protein expression. 70

A B C 71

Tabic 4. Shedding of swine influenza virus in vaccinated and nonvaccinated pigs after challenge witii virulent virus

Vaccine titer (log lO) oronasal swabs Vaccine given Pig No. and route Days shed Day 5 titer Day 6 titer Day 7 titer

77 None 7 4.7 1.0 0.5 79 7 5.5 3.7 0.5 91 7 3.7 1.0 1.7 2239 6 2.5 1.7 0.0 80 MVA,IM 6 4.0 0.5 0.0 81 7 2.7 1.0 0.5 82 6 2.5 0.5 0.0 83 6 42 1.7 0.0 84 MVA,IN 6 2.7 1.5 0.0 85 6 3.2 0.7 0.0 86 6 3.7 0.5 0.0 87 6 32 0.5 0.0 78 MVA(R), IM 6 22 1.2 0.0 88 6 0.7 0.5 0.0 89 5 0.7 0.0 0.0 90 5 0.5 0.0 0.0 93 MVA(R), IN* 4 0.0 0.0 0.0 94 5 0.5 0.0 0.0 95 6 2.5 1.2 0.0

* The 4th pig in the MVA(R), IN group died prior to challenge from trauma unrelated to vaccine or challenge. Table 5. Serological response in vaccinated and nonvaccinated pigs to challenge with virulent influenza virus.

HAI titers vs PR/8 * HAI titers vs IN/88 Vaccine given 14 dpv 14 dpv 7 dpc 7 dpc cELISA result 7 dpc AGID result 7 dpc vs Pig No. and route #1 #2 vs Type A NPt Type A NP, M protein ^

77 None n n n 10 n n 79 n n n 40 n wl 91 n n n 20 s wl 2239 n n n 80 n 1

80 MVA, IM n n n 40 n n 81 n n n 80 n 2 82 n n n 40 s wl 83 n n n 20 s wl

84 MVA, IN n n n 10 n n 85 n n n 80 n n 86 n n n 20 n wl 87 n n n 20 + 1

78 MVA (R), IM 10 160 160 20 s 3 88 <10 320 320 40 + 2 89 40 320 320 20 + 3 90 n 320 <320 10 + 3

93 MVA (R), IN n 80 80 40 + 4 94 n 40 40 40 + 3 95 n 40 40 <10 s 2

* Tilers are expressed as the reciprocal of the last dilution which inhibits 4 HA units of virus. dpv #1, dpv ^2 = days post vaccination #1 and #2, dpc = days post challenge; n = negative. t For cELlSA, n = negative (0.0); s = suspect (0.5); + = positive (1.0). t For AGID, n = negative (0.0); wl = very weak positive (0.5); 1,2 = weak positive (1.0,2.0); 3 = positive (3.0); 4 = strong positive (4.0). 73

Table 6. Virus shedding by vaccinated and nonvaccinated pigs after challenge with swine influenza virus. Group Route of Mean days vaccination virus shedding*

Controls — 6.75 MVA IM 6.25 MVA EST 6.00 MVA(R) IM 5.50 MVA(R) IN 5.00

* There was a significant difference in vims shedding between pigs vaccinated with the recombinant and the nonvaccinated controls, following challenge with SW/IN/88: Recombinant IM vs Control (p <0.05), Recombinant IN vs Control (p <0.05), Recombinant IM vs Parent IM (p >0.08), Recombinant IN vs Parent IN (p <0.05) and Recombinant IM vs Recombinant IN (p >0.25). MVA (R) = MVA-HA-NP. 74

Table 7. Effect of immunization with recombinant vaccinia vector against SIV challenge.

Route of Virus titers at DPC Group vaccination Day 2 Day 3 Day 4

Control — 3.56 3.56 4.88* MVA (R) - IM IM 4.13 2.50 3.56 MVA (R) - IN IN 4.42 2-83 1.83*

*There was no significant difference ia virus titers among groups until day 4 post challenge, when the animals vaccinated intranasally demonstrate a significantly lower titer of virus shedding than the controls (p <0.05). DPC = days post challenge; MVA (R) = MVA-HA-NP. 75

Table 8. Comparison of virus shedding in vaccinated and control pigs 5 days after challenge with swine influenza virus.

Group Route of Mean EID50* vaccination of virus shed

Controls - •- 4.25 MVA IM 3.38 MVA EST 3.25 MVA(R) IM 1.06 MVA(R) IN 1.00

*By day 5 post challenge, both groups vaccinated with the recombinant had significantly lower amounts of virus shed: Recombinant IM vs Control (p <0.05)* Recombinant IN vs Control (p <0.05)* Recombinant IM vs Parent (p <0.05)* Recombinant IN vs Parent (p <0.05)* Recombinant IM vs Recombinant IN (p >0.50) EID50 = Mean egg infectious dose; MVA (R) = MVA-HA-NP recombinant. 76

Table 9. Humoral response in pigs to HA and MP after challenge with virulent swine influenza virus.

HAI GMT Mean cEIJS A OD at Mean AGED values at Group 7 dpc vs IN/88 7 dpc vs type A NP 7 dpc vs type A NP, M protein

Controls 28 0.13 (N to S) 0.50 (VWP) MVA-IM 40 0.25 (N to S) 0.75 (VWP to Weak) MVA-IN 24 0.25 (N to S) 0.38G^^egtoVWP) MVA(R)-IM 20 0.88 (S) 2.75 (Weak to Pos) MVA (R)-IN 20 0.83 (S) 3.00 (Pos)

<10 Gog = 0) Neg = 0.0 Neg = 0.0 10 Gog =1) Sus = 0.5 Very Weak Pos = 0.5 20 Gog = 2) Pos = 1.0 Weak Pos = 1.0 or 2.0 40 Gog = 3) Pos = 3.0 80 Gog = 4) Strong Pos = 4.0

1. HAI No significant group differences (p>0.75)

2.ELISA Recombinant IM vs Control (p <0.05)* Recombinant IN vs Control (p <0.05)* Recombinant EM vs Parent IM (p <0.05)* Recombinant IN vs Parent IN (p <0.05)* Recombinant IM vs Recombinant IN (p >0.50) 3.AGID Recombinant IM vs Control (p <0.05)* Recombinant IN vs Control (p <0.05)* Recombinant IM vs Parent IM (p <0.05)* Recombinant IN vs Parent EN (p <0.05)* Recombinant IM vs Recombinant IN (p>0.50)

Using an Analysis of Variance and then the Least Significant Difference method (LSD) to compare means, both groups vaccinated with the MVA recombinant had significantly greater antibody titer to NP, but not HA, when compared to all other groups. GMT = geometric mean titer; dpc = days post challenge. 77

2. RECOMBINANT VACCINIA VIRUS CONTAINING HEMAGGLUTININ AND NUCLEOPROTEIN GENES FROM A PORCINE STRAIN OF INFLUENZA VIRUS PROTECTS PIGS AGAINST SWINE INFLUENZA VIRUS

A paper to be submitted to the American Journal of Veterinary Research

Patricia L. Foley, MA, DVM; Bruce H. Janke, DVM, PhD; lone R. Stoll; Steve Hanson; Prem S. Paul, BVSc, PhD

Structured Abstract

Objective - To construct and evaluate eflBcacy of an attenuated recombinant vaccinia virus expressing hemagglutinin (HA) and nucleoprotein (NP) genes from a field isolate of SAvine influenza virus (SIV).

Animals - 32 caesarean-derived pigs given only SIV antibody-free colostrum post farrowing were used for immunization and evaluation of eflBcacy of the recombinant vaccine.

Procedure - The HA and NP genes of SIV were amplified by reverse transcription polymerase chain reaction (RT-PCR) and cloned. These genes were sequenced and subcloned into a transfection vector that facilitated insertion of the SIV sequences into the genome of the modified vaccinia virus Ankara (MVA) strain of vaccinia virus. Plaques were screened for HA and NP protein expression, purified, and expanded. Thirty pigs of mixed sex were placed into five groups of six pigs each. An additional two pigs served as nonvaccinated, nonchallenged controls. Two groups of six were vaccinated with the 78

MVA/SrV recombinant, one intramuscularly (IM) and the other intranasally (IN), at 20-22 days of age. Another two groups of six were vaccinated with an MVA recombinant containing HA and NP genes from the human strain A/PR/8/34 (PR8), again one IM and the other EST, at the same dosage. Those four groups were reinoculated at 34-36 days of age. At

49-52 days of age, all pigs, including a fifth group of 6 nonvaccinated controls, were oronasally challenged with the homologous SIV strain, using a nebulizer. The pigs were monitored and nasal swabs collected until day 5 postchallenge, at which time all 32 pigs were euthanized and postmortem lesions examined grossly, histologically, and by immunohistochemistry for the presence of SIV.

Results - Sequence analysis determined the close similarity of this HlNl strain with a standard 'classic' strain of SFV, as well as its relative distance from the PR8 strain of influenza. The MVA/SFV recombinant expressed well its foreign proteins, grew to high titer in pure culture, and induced an immune response in vaccinates. Following vaccination and challenge, significant differences were seen between vaccinates and controls in terms of clinical scores, days and titers of virus shedding, and lung lesions.

Conclusions - The vaccination/challenge study indicated the ability of the MVA/SFV vaccine, especially given IM, to protect pigs against clinical signs, viral shedding, and gross and histological lesions tj^jically associated with swine influenza. Clinical Relevance - Awareness of the relative ease with which strains of influenza virus can

reassert have constramed attempts to develop traditional modified live virus vaccines. The

recombinant MVA/SIV vaccine provides a modified live virus alternative, inducing humoral and cell-mediated immune responses, against SIV. That it appears to work well given IM is advantageous to its application in the field.

Introduction

Many pathogenic viruses, against which vaccines would be desirable, are simply too

proficient at producing disease to be administered in a live product. In addition, even when a virus strain does not itself produce disease, it may have characteristics that, when recombined with those of another strain encountered in the field, result in a far worse disease outbreak.

Such is the case with influenza viruses, which have a segmented genome capable of reassortment. If an attenuated vaccine str^n, with certain gene segments that give it the ability to replicate well in a given host species, should exchange segments with another virulent strain that does not replicate well in that species, the outcome may be a new strain that is both virulent and well-adapted to the new host. That was the fear that characterized the 1997 Hong Kong outbreak in humans of an avian strain of influenza Fortunately,

although the avian strain was able to colonize certain humans, it did not replicate well enough

in that host to cause an epidemic in what would have been a completely susceptible,

immunologically naive population.

Among the veterinary influenza products, there are eight avian vaccines manufactured by one company, one swine vaccine, and five equine vaccines produced by six different 80 companies They are all killed virus products, as are the human influenza vaccines used in annual vaccination programs. Human vaccines are changed annually to reflect the newly emerging strains and frequently contain 3 distinct hemagglutinin antigens from diverse strains to maximize protection against disease

One problem with inactivated influenza virus vaccines is that the immunity generated is only partial In the presence of a strong adjuvant, antigens can stimulate B-cells and induce a good humoral response. However, there is little cell-mediated immunity generated by a killed product, and this can mean the difference between disease and protection or, at least, a more rapid recovery from disease Also, the immunity provided by a killed product can be relatively short-livedThe potential advantages of a recombinant vaccine are that it may express protective immunogens against even the most dangerous of viruses in a safe vector, provide both humoral and cell-mediated immunity and extend the duration of that protection beyond the time provided by a killed product.

Vaccinia virus, a member of the Orthopoxviruses," is known to be a strong inducer of humoral and cell-mediated immunity Vaccinia virus has been used as an experimental vector in several species and in accidental human exposure with considerable documentation of its ability to induce immunity to several diseases. The first reports of its use as a vector for the delivery of immunogens surfaced in the early 1980's. By 1990, there were numerous reports in the literature regarding its successful expression of the influenza virus hemagglutinin or nucleoprotein the hepatitis B virus surface antigen the glycoprotein D from herpes simplex virus " the rabies virus glycoprotein envelope a glycoprotein of bovine leukaemia virus and other proteins. Several studies demonstrated 81 its immunogenicity in humans Insertion of more than one gene was also accomplished, such as one recombinant expressing both the influenza virus hemagglutinin and the herpes simplex virus thymidine kinase gene or another expressing the hepatitis B virus surface antigen, the herpes simplex virus glycoprotein D, and the influenza virus hemagglutinin

This raised the possibility of a single vector expressing immunogens from multiple pathogens, thereby providing protection against all of them with a single vaccine.

However, as promising as vaccinia \'irus seemed to be for certain applications, it had several drawbacks which limited its potential for general use. During the smallpox eradication effort, it was known to cause unwelcome side effects. Apart from the localized irritation induced by the inoculation, there were more severe complications among the immunocompromised. It was estimated that one in every million vaccinations resulted in death For this reason, following the end of smallpox vaccination, vaccinia virus was no longer used clinically in the U. S. or the rest of the world. Once well-exposed to this virus, following decades of little to no use, the general population has now become quite susceptible, as people less than 30 years of age have no immunity. Unfortunately, this risk of human exposure greatly curtails the usefulness of vaccinia as a vector. Even if a vaccine were designed for some one species in the veterinary market, because vaccinia has such a wide host range, there exists the potential for inadvertent inter-species spread.

To circumvent the problems associated with vaccinia virus, yet retain its advantages, researchers have investigated strains of vaccinia virus that are attenuated, either by nature or design. One such strain is the modified vaccinia Ankara (MVA), for which there have been numerous studies jj. originally developed from the vaccinia virus Ankara 82 strain as a safe alternative for smallpox vaccination, and has been used without significant side-effects in over 120,000 people, including young children and the elderly, for immunization against smallpox. After approximately 570 passages in primary chick embryo fibroblasts (CEF), it has lost its ability to replicate or at least replicate well in numerous mammalian cell lines. It contains six major deletions that prevent virus assembly in most all mammalian cells tested but leave gene expression, both early and late, relatively unimpaired.

The exact nature of this host restriction is not really understood. Thus far, four orthopox virus host range genes have been documented. These are the CHOhr C7L KIL and E3L genes. Of these, only the fiinction of the E3L gene, which expresses an RNA binding protein is known. Compared to its parental strain, MVA has deletions that consist of about 15% (30,000 base pairs) of its former genome, including most of the KIL gene

Interestingly, in one study, replacement of the KIL gene in MVA removed only the host restriction in RX13 cells This suggests that there are multiple, cumulative genetic defects in MVA replication. If so, as seems likely, the probability of spontaneous reversion to a wild type host range is quite low, which increases the safety of MVA as a vaccine vector.

In this study, the genes for two immunogenic proteins of the swine influenza virus (SIV) were inserted into the MVA strain. The construct was then evaluated for protein expression and ability to protect pigs against virulent SIV. We report that this recombinant SIV vaccine was a safe and efiBcacious deterrent to disease. 83

Materials and Methods

Construction of the recombinant - Genomic RNA from the ISU SIV isolate was purified from infected allantoic fluid using the Purescript RNA isolation kit (Centra), as detailed in Figure 1. HA- and NP-specific primers were used in a Titan One-Tube RT-PCR reaction (Boehringer Mannheim) to generate first strand cDNA, then amplified to double- stranded DNA. The RT-PCR products were cloned into the CloneAmp pAMPl System vector (GibcoBRL) for sequencing and subcloning into the JS5 plasmid vector. JS5 contains the same double promoters in opposite orientation for dual-gene cloning but not the flanking

MVA sequences as found in plasmid vector G06 (both vectors courtesy of B. Moss,

Laboratory of Viral Diseases, NLAED, NIH). It also has more convenient restriction sites for gene insertion. Following proper insertion of the two SIV genes, the cassette was removed from JS5 and inserted into G06. Following PCR screening of transformed colonies, using

HA- and NP-specific primers, plasmid DNA was cut with various en2ymes to determine correct orientation (figures 2a, 2b, 2c). DNA from 4 clones was used in the presence of

Lipofectamine Plus reagent (Gibco BRL) to transfect MVA-infected cells (figure 3). Cells were passed and positive plaques purified at least six times (figure 4), as previously described^®. One plaque was selected for expansion and designated MVA/SIV.

Cells and Viruses - Vaccinia virus strains MVA and MVA/PR8 were generously provided by B. Moss, Laboratory for Viral Diseases (LVD), National Institute of Allergy and

Infectious Diseases (NIAID), National Institutes of Health (NIH). MVA/SIV was recently constructed in our laboratory, as described above, from an Iowa field strain of SIV isolated 84 by the Iowa State University Veterinary Diagnostic Laboratory (courtesy of B. Janke).

MVA, MVA/PR8, and MVA/SIV were grown at 37°C and 5% CO2 in the second passage of chicken embryo fibroblast cells, using M199 medium with F-IO nutrient mixture (M199/F-

10), supplemented with 0.15% bactotryptose phosphate broth, 0.09% Na bicarbonate, 1%

200 mM L-glutamine, 25 U/ml penicillin G potassium, 75 U/ml streptomycin sulfate, 0.1% gentocin, and 5% FBS.

A I960's SrV field isolate, provided by the Diagnostic Virology Laboratory (DVL),

National Veterinary Services Laboratories (NVSL), and the ISU field isolate were propagated in the allantoic cavity of 10-day-old embryonated chicken eggs at 37°C for 72 hours. The mean egg infectious dose (EID50) and hemagglutination (HA) units of these SIV lots were determined. In addition, a 24-hr-old monolayer of swine testicular (ST) cells, seeded at 2 x 10^ cells/ml, and grown at 37°C, in 5% CO2, using Minimum Essential Medium

(MEM) with Earle's salts (Gibco cat. no. 41500-018), supplemented with 0.22% Na bicarbonate, 0.5% edamine, 1% 200 mM L-glutamine, 25 U/ml penicillin G potassium, 75

U/ml streptomycin sulfate, 0.1% gentocin, 1% sodium pyruvate, and 5% FBS, was used for

SrV replication. At 24 hours after seeding, medium was decanted and the ST monolayer inoculated with virus in sufficient medium to cover the monolayer. Following virus adsorption for one hour at 37°C, fi^esh medium was added and the cultures incubated at 37°C in 5% COi.

Titration of viruses - For MVA and MVA recombinant virus titrations, 24-hour old

CEF cells, seeded at 8 x 10^ cells/ml in 60 millimeter (mm) tissue culture plates under 85 conditions cited above, were prepared. Dilutions of virus were made in M199-F10 media containing 2% FBS. Growth medium on the 60 mm plates was decanted and 0.2 ml per dilution added. Plates were held at 37°C, 5% COj, for one hour, with rocking every 15 mins.

The inoculum was then aspirated and the plates refed with 4 mis of 2X MEM containing 10%

FBS, 1% 200 mM L-glutamine, 25 U/ml penicillin G potassium, 75 U/ml streptomycin sulfate, and 0.1% gentocin mixed in equal parts with 1.1% melted agar. The mixture was allowed to solidify in the plates, which were then incubated at 37°C, 5% CO,, for 2 to 3 days.

The agar was removed by washing gently with 0.01 M PBS, pH 7.2, and the plates fixed.

MVA recombinant-inoculated plates were fixed in 1:1 acetonermethanol, reacted with primary antibody, then anti-mouse or anti-swine peroxidase-labeled conjugate. The substrate used for these assays was metal enhanced DAB (Pierce, Rockford, EL). Recombinant virus titers were determined using insert-reactive anti-HlNl SIV polyclonal antisera and anti-PRS

HA and NP monoclonal antibodies (courtesy of J. Yewdell, NIAID, NTH) as primary antibody, and expressed as plaque-forming units (PFU). MVA titers were determined using either cytopathic effect (CPE) or immunofluorescent assay (IFA). For the latter procedure, plates were fixed in 80% acetone. The primary antibody was a rabbit-origin anti-vaccinia virus polyclonal antisera, reacted with an anti-rabbit FITC-conjugate. The titers were expressed as fluorescent focal units (FFU). This EFA procedure was also used on the MVA recombinant virus-infected cultures, subsequent to the DAB immunoplaque assay, to determine the titer of non-expressing plaques.

For SIV titration on the ST cell line, 96-well plates were inoculated with 50 ul virus dilution per well, centrifiiged at room temperature for 2 hours at 400xg, incubated at 37°C in 86

5% COj, then read by CPE or IF A, using anti-HlNl SIV polyclonal antisera, 5-7 days postinoculation.

Immunoassays - Plaque assays to evaluate recombinant protein expression were performed in a maimer similar to virus titrations, described above. For plaque purification, however, a live immunostaining procedure was employed, i.e. virus-infected plates were not fixed. All other steps were the same.

Vaccination/Challenge - At 20-22 days of age, four groups of six caesarean-derived pigs, that were fed colostrum (Colostrix, Struve Labs) negative for SIV antibody, were vaccinated with 10® ° PFU of the plaque-purified MVA/PR8 or MVA/SIV recombinant either intramuscularly (TM) or intranasally (IN). Immunization was repeated at 34-36 days of age.

At 49-52 days of age, these twenty-four pigs and six unvaccinated controls were challenged oronasally with approximately 2 mis of the homologous ISU field strain (10®-® EIDso/ml). A

Satum Nebulizer (Dynax), with an attached plastic face mask held over the animals' faces for

5 minutes, w^as used to deliver virus to the lower airways of the respiratory tract. At 5 days postchallenge, the 30 pigs plus 2 nonvaccinated, nonchallenged control pigs were euthanized and lung tissue examined.

Virus isolation and titration - Nasal swabs were collected from all pigs at 2, 3, 4, and 5 days post-challenge and titrated in embryonated chicken eggs to determine levels of virus shedding. The nasal swab samples were placed in tubes containing 2 mis Eagle's MEM 87 supplemented with 75 U/mi penicillin G potassium, 225 U/ml streptomycin sulfate, 0.1% gentocin, and 1.5% amphotericin B. The tubes were frozen, then thawed, vortex-mixed, and centrifliged at 400xg for 10 minutes. Ten-day-old embryonated eggs were inoculated in the allantoic cavity with 0.1 ml of the resulting supernatant and incubated at 37° for 72 hr. The harvested allantoic fluids were tested for HA activity. Eggs were inoculated in duplicate to determine the presence of virus. Samples found to be positive for virus were then titrated.

For titration, four eggs were inoculated with each dilution and the results tabulated using the method of Spearman-Karber

Serological assays - Serum samples were harvested prior to first vaccination, second vaccination, challenge, and euthanasia, then analyzed by hemagglutination inhibition (HAI) against the two strains of influenza virus, strains PR8 and ISU, and by serum neutralization

(SN) of MVA and SW, using the constant virus-varying serum method.

HA and HAI tests were performed in 96-well microtiter plates, using 0.5% chicken erythrocytes. For the SIV and PR8 HAI assay, sera were pretreated using 10% kaolin and

5% washed chicken erythrocytes, then evaluated at 1:10 or greater dilutions against standardized live virus antigen. For these assays, four HA units of each virus were used to determine serum HAI titers.

The MVA SN procedure used 96-well plates containing 24-hour-old monolayers of CEF cells, inoculated with 50 ul of a 1:1 mixture of virus and 5-fold dilutions of serum, previously incubated at 37°C for 1 hour. The inoculum was decanted afl;er 1 hour, replaced with 200 ul of M199/F-10 media, supplemented as above but with only 0.5% FES, and read at 5-7 days 88 by CPE. SN titers were determined by the method of Reed-Muench

The SrV SN procedure is the same as that used for the MVA SN assay, except for the use of the ST cell line, a different maintenance medium (MEM and Earle's salts with no

FBS), and centrifugation, as described above.

Post mortem examination - Following euthanasia, lungs were examined grossly and histologically for lesions. Tissues were fixed in 10% formalin, processed to paraflBn blocks, and mounted on poly-L lysine coated glass slides. An immunohistochemical procedure, using a monoclonal antibody produced in mice against influenza viral nucleoprotein, was employed to determine the presence of viral antigen in lung tissue.

Statistical analysis - Data was evaluated using an analysis of variance and then the

Least Significant Difference method to compare means. To determine geometric mean titer

(GMT), logio of each titer was summated, then averaged and the antilog determined.

Because negative titers are '0', 1 was added to all numbers, then subtracted later fi-om the averaged GMT.

Results

Vaccination/Challenge - Different clinical signs were assigned a numerical value (table la) and scored (table lb), with the challenged nonvaccinates having far more signs than the vaccinates, including marked anorexia (not scored). The MVA/SIV vaccinates especially seemed unaffected by the challenge. Upon necropsy, they had very little to no gross or 89 histopathologicai lung lesions, and little to no virus present in the lung as determined by immunohistochemistry (table 2).

To summarize the gross lesions, they were most severe in the challenged, nonvaccinated pigs, with total lung involvement of 2-30%, ahnost entirely affecting cranial and middle lung lobes. There were lesions of similar extent and distribution in MVA/PR8 IM and IN pigs with <10% total lung involvement in 5 of 6 pigs and 15% in 1 of 6 in each group. One pig in the MVA/PR8 IM group had almost no lesions; one pig in the MVA/PR8 EST group had lesions nearly as extensive as in some non-vaccinates. There were only minimal lesions in the tips of the middle lobes in 2 of 6 IVIVA/SIV IN pigs, with no lesions at all in the other 4 of 6

MVA/SrV IN pigs or in the MVA/SIV IM pigs.

Regarding the microscopic lesions, those typical of SIV infection were found in the challenged nonvaccinates, with active necrosis of bronchiolar epithelium and proliferative lesions of repair continuing 5 days after infection, particularly in the smaller bronchioles.

There were lesions of similar character but much more focal in MVA/PR8 IN pigs. Lesions were also more characteristic of the repair stage, with minimal active necrosis in these pigs.

Lesions were very mild in the MVA/PR8 IM pigs. There were similar mild lesions in 2

MVA/SIV IN pigs but slightly more severe in one. There were no lesions in the other 3

MVA/SrV IN pigs, as well as no lesions in the MVA/SIV IM pigs, with one small exception

(2 small foci in a cranial lobe). In general, although gross lesions in MVA/PR8 IM and IN pigs appeared similar, microscopically, damage in the IM-vaccinated pigs was less severe.

Likewise, microscopic lesions were almost non-existent in MVA/SIV IM pigs but were focally present, though mild, in some MVA/SIV IN pigs. 90

Analysis by immunohistochemistry detected most infected cells in the challenged nonvaccinates, which correlates with the presence of active ongoing necrosis of epithelium in these pigs. Virus was also still present in the alveoli in these pigs. There was a dramatic reduction in virus in ail vaccinated pigs. Although there were more lesions induced in pigs vaccinated with MVA/PR8 than in pigs vaccinated with MVA/SIV by the corresponding route, virus is apparently cleared more rapidly in vaccinated pigs regardless of the MVA constructs employed.

There were no clinical signs of disease nor lesions of any kind in the two nonvaccinated, nonchallenged control pigs.

Virus isolations and titrations - Challenged nonvaccinates continued to shed virus through day 5 after challenge. In contrast, vaccinates, with the exception of the MVA/PR8

IM group, shed on fewer days and at much lower titer (table 3).

Serological assays - Serum HAI titers were higher in MVA/PR8 IM vaccinates than in the MVA/PR8 IN group, but this was not so evident in the MVA/SIV groups. Detectable titers did not appear in the MVA/SIV vaccinates until 14 days after the second vaccination.

These were boosted by the SIV challenge as early as 5 days postchallenge, unlike those of the

MVA/PR8 groups (table 4, figure 5). The SN titers for both MVA and SIV were very low prechallenge. At 5 days postchallenge, all vaccinated groups had some SIV SN titer developing (table 5). 91

Statistical analysis - The cumulative clinical sign score for the challenged nonvaccinated

pigs was significantly greater than all other group scores (p<0.05), but the scores for the

vaccinated groups were not statistically different from each other (table lb). The challenged

nonvaccinated pigs had significantly more virus isolations than all groups (table 3), except those vaccinated IM with the MVA/PR8 recombinant. The MVA/PR8 IM and EN groups

had a significantly greater number of isolations than the respective MVA/SIV IM and IN

groups, but the two MVA/PR8 groups, as well as the two MVA/SIV groups, were not

different from each other. Regarding titers of virus shed, the nonvaccinated controls had significantly greater titers than the MVA/SIV IN group on all 4 days of swabbing, the

MVA/SrV IM and the MVA/PR8 EST groups on days 3-5, and the MVA/PR8 IM group on days 4 and 5 (data not shown). Analysis of the HAI titers indicates that, on the day before and 5 days postchallenge, titers of pigs vaccinated with MVA/PR8 IM are significantly greater than those for the MVA/PR8 IN group (table 4). However, there is no significant difference in titers between the two MVA/SIV groups for those days. For the serum neutralization titers, on day 5 postchallenge, each of the vaccinated pig groups had significantly higher titers than the nonvaccinated controls, but there were no significant differences between the vaccinated groups (table 5).

Discussion

The pathology induced by SFV occurs throughout the respiratory tract and consists of acute inflammation, edema, and bronchiolar epithelial necrosis Lung lesions tend to be bilaterally distributed, predominantly in the cranial and middle lobes. Generally, most pigs 92 recover but resolution of lesions may take up to a month. The MVA/SIV recombinant was eflBcacious in preventing these typical signs and lesions of SIV.

In recent years, however, there have been periodic occurrences of 'atypical' SIV outbreaks, leading to speculation that the relatively stable antigenic profile of SIV may be changing. Reports have appeared in the literature since 1992, indicating the occurrence of

SrV either associated with unusual signs or exhibiting more virulence than expected. First, there was a report fi'om Quebec regarding an HlNl variant producing proliferative and necrotizing pneumonia in pigs, with some signs very similar to those of the PRRS virus with more point mutations and diversity than generally seen in North American SIV isolates Another novel isolate, from a severely affected herd, was designated

A/Sw/Nebraska/1/92 It induced persistent, high fevers (up to 42°C) but not much respiratory disease. Given the high degree of conserved sequences in typical SIV strains isolated in the U. S., it was surprising that the most closely related reference HlNl strain had only 94% identity at the nucleotide level and 96% at the amino acid level to this SIV isolate.

Nonetheless, it was closest genetically to 'classical' HlNl SIV than to avian or human HlNl viruses. In England, an HlNl strain antigenically distinguishable from classic SIV and

European avianlike HlNl viruses caused a sudden increase in SIV cases, but still exhibited the usual clinical signs of coughing, snee2dng, and anorexia However, upon experimental infection, this strain produced a more severe interstitial pneumonia and hemorrhagic lymph nodes.

The significance of the genetic diversity represented by these strains is as yet undetermined. It may be that there are no 'atypical' SFV strains, merely a greater degree of 93 potential antigenic diversity among field strains than previously detected. It is known, however, that subtype HlNl influenza viruses have been continuously circulating in U. S. pigs for over 60 years. It is believed that the great pandemic of "Spanish flu" in 1918/19, the worst in history, killing at least 20 million people worldwide, was either caused by a swine virus or by a human strain that entered the pig population at that time In 1997, RNA from a casualty was extracted from formalin-fixed, parafiBn-embedded tissue and sequenced All sequences determined were very similar to those of classic HlNI SFV, suggesting that human and swine strains share a common avian ancestor, existing some time before 1918. The first isolation of SFV did not occur until 1930, however, and the 'classical' HlNl swine virus recovered (A/Swine/Iowa/15/30) is still much like the majority of SIV found circulating in

U. S. pigs today. The ISU isolate from which HA and NP genes were cloned is representative of this subtype, as determined by sequencing. Given the results of our study, the MVA/SIV recombinant should provide protection against these strains.

Northern Europe saw its first isolate of SFV in 1978/79, and although an HlNl virus, its

HI was similar to the avian HI but distinct from both human and swine HI Since then, there have been instances where an avian virus has been able to cross species and infect the pig population, as with SwGer/81, and cases where reassortment between avian and classical

SrV has occurred, as with SwHK/82 . In addition, there is a human virus-like H3N2 subtype that has been isolated on occasion in European pigs since 1980, perhaps as a result of the 1968 antigenic shift and an ability of the H3N2 to persist in pigs even when not circulating in the human population It is interesting that a serological survey conducted in

1988-1989 in U. S. pigs found evidence of H3 viruses antigenically similar to the then current 94 human H3 strains, at about 1.1% average prevalence In addition, a serological survey during 1976-1977 detected an incidence of 1.4% for H3N2 infections^. Moreover, in that study, isolation from one herd of a virus antigenically similar to a human H3N2 strain was reported- However, complete sequencing to determine whether the isolate was of human origin was not performed. So, no H3 human strain has been confirmed as present in U. S. pigs. Very recently, the National Veterinary Services Laboratories identified an influenza virus subtype H3N2 isolate from a swine breeding herd in North Carolina (interlaboratory communication). Studies are currently underway to determine species of origin and other characteristics of the virus.

So, at the very least, there are three HA subtypes circulating in pigs at present, classic

SrV HI, avian \arus-like HI, and human virus-like H3. They have been found in various permutations of SIV gene segments. One, an H3N1 strain, appeared to be a combination of the classic SFV and the human virus-like H3N2 found in swine "•*. Another, an H1N2 isolate believed to be from a human HlNl and the svme-adapted H3N2, caused clinical disease in pigs Still another represented a reassortment between human and avian strains in symptomatic Italian pigs, providing the first proof that pigs can act as 'mixing vessels' for human and avian viruses The critical role that pigs can play in pandemics was underscored with the discovery that children in the Netherlands were sick from avian-human influenza virus generated in pigs, transmitted pig-to-person, and person-to-person Normally, avian strains do not replicate in humans, and human strains do not replicate in birds. This is a fijnction of their specific sialyloligosaccharide receptors on the surface of epithelial cells of the upper respiratory tract. In a previous study, it was determined that, of 38 avian influenza 95 strains, fiiliy 31 were successfully transmitted to swine Every HA subtype (of 14 tested) had at least one strain that grew as well as a swine or human virus. Since then, it has been determined that pigs, in fact, have both avian- and human-specific viral receptors present in their upper respiratory tract and that some avian strains, with continued replication, acquire the ability to recognize human receptors as they become sAvine-adapted

Taken together, these data delineate the danger to humans of a pig population unprotected against influenza. If an avian virus with a non-human-type HA is introduced into pigs, then reassorts with a human strain, a pandemic among complete susceptibles would occur. Although direct interspecies spread fi-om bird to human can happen, as was seen with the 1997 Hong Kong H5N1 cases virus under those circumstances may not readily adapt to its new host and relatively few may be affected. The dangers may be greater in the former scenario.

In summary, interspecies transmission is known to occur and the pig, able to be infected by either avian or human strains can serve as a 'mixing vessel', wherein gene segments from different strains reassort to produce new viruses. Clearly, a safe, live vaccine vector able to express multiple genes, that could be given frequently to boost or provide new immunity, would be helpful. The recombinant MVA'SIV vaccine developed and evaluated in this study holds promise as a safe, effective means to protect pigs and people against typical U. S. strains of HlNl SFV. Even when the HA gene of the insert is heterotypic to an infecting strain, as was the case for the MVA/PR8 construct, immunity generated against the type A

MP can help reduce clinical signs and clear virus more rapidly from vaccinated pigs.

MVA may also be suitable as a vector for delivery of immunogens from other strains of 96 influenza. For repeated use of MVA in this capacity, it will be necessary to demonstrate that preexisting immunity to MVA or expressed proteins of one recombinant vaccine does not inhibit the immune response to a second recombinant vaccine.

References

1. Altenburger W, Suter C-P, Altenburger J. Partial deletion of the human host range gene in the attenuated vaccinia virus MVA. Arch Virol 1989;105:15-27.

2. Baxby D. The origins of vaccinia. Tenner's Smallpox Vaccine. London: Heinemaim; 1981. p 179-96.

3. Beattie E, Kauffman EB, Martinez H, Perkus ME, Jacobs BI, Paoletti E, Tartaglia J. Host-range restriction of vaccinia virus E3L-specific deletion mutants. Virus Genes 1996;12:89-94.

4. Bender BS, Rowe CA, Taylor SF, Wyatt LS, Moss B, Small Jr. PA. Oral immunization with a replication-deficient recombinant vaccinia virus protects mice against influenza. J Virol 1996 Sep;70(9);6418-24.

5. Blancou J, Kieny M-P, Lathe R, Lecocq JP, Pastoret PP, .Soulebot JP, Desmettres P. Oral vaccination of the fox against rabies using a live recombinant vaccinia virus. Nature 1986;332;373.

6. Brown IH, Chakraverty P, Harris PA, Alexander DJ. Disease outbreaks in pigs in Great Britain due to an influenza A virus of H1N2 serotype. Vet Rec 1995;136:328-9.

7. Brown Di, Done SH, Spencer YI, Cooley WA, Harris PA, Alexander DJ. Pathogenicity of a swine influenza HlNl virus antigenically distinguishable from classical and European strains. Vet Rec 1993;132:598-602.

8. Buller RML, Palumbo GJ. Safety and attenuation of vaccinia virus. Birms MM, Smith GL, eds. Recombinant Poxviruses. Boca Raton: CRC Press; 1992. p 235-67.

9. Carroll MW, Moss B. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 1997;238:19S-211. 97

10. Carroll MW, Overwijk WW, Chamberlain RS, Rosenberg SA, Moss B, Restifo MP. Highly attenuated modified vaccinia virus Ankara (MVA) as an effective recombinant vector: a murine tumor model. Vaccine 1997;15(4):387-94.

11. Castrucci MR, Donatelli I, Sidoli L, Barigazzi G, Kawaoka Y, Webster RG. Genetic reassortment between avian and human influenza A viruses in Italian pigs. Virology 1993; 193:503-6.

12. Chambers TM, Hinshaw VS, Kawaoka Y, Easterday BC, Webster RG. Influenza viral infection of swine in the United States 1988-89. Arch Virol 1991;116:261-5.

13 Chang H-W, Uribe LH, Jacobs BL. Rescue of vaccinia virus lacking the E3L gene by mutants of E3L. J Virol 1995;69:6605-8.

14. Chang H-W, Watson JC, Jacobs BL. The E3L gene of vaccinia virus encodes an inhibitor of the interferon-induced, double-stranded RNA-dependent protein kinase. Proc Natl Acad Sci U S A 1992 Jun;89;4825-9.

15. Chelyapov NV, Antonova TP, Yanova NN, Chemos VT. Antigenic properties of vaccinia virus and of the virus recombinant strains expressing heterologous genes. Acta Virol (Praha) 1988;32:409-16.

16. Claas ECJ, Kawaoka Y, de Jong JC, Masurel N, Webster RG. Infection of children with avian-human reassortant influenza virus from pigs in Europe. Virology 1994;204;453-7.

17. Claas ECJ, Osterhaus ADME, van Beek R, De Jong JC, Rimmelzwaan GF, Senne DA, Krauss S, Shortridge ECF, Webster RG. Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus. Lancet 1998 Feb;351:472-7.

18. Claas ECJ, de Jong JC, van Beek R, Rimmelzwaan GF, Osterhaus ADME. Human influenza virus A/Hong Kong/156/97 (HlNl) infection. Vaccine 1998;16(9/10):977-8.

19, Cooney EL, McElrath MJ, Corey L, Hu SL and others. Enhanced immunity to HIV envelope elicited by a combined vaccine regimen consisting of priming with a vaccinia recombinant expressing HTV envelope and boosting with gpl60 protein. Proc Natl Acad Sci U S A 1993 Mar;90(5):1882-6.

20. Couch RB, Keitel WA, Cate TR, Quarles JA, Taber LA, Glezen WP. Prevention of influenza virus infections by current inactivated influenza virus vaccines. Brown LE, Hampson AW, Webster RG, eds. Options for the control of influenza HE. Amsterdam: Elsevier; 1996. p 97-106. 98

21. Coupar BEH, Andrew ME, Boyle DB. A general method for the construction of recombinant vaccinia viruses expressing multiple foreign genes. Gene 1988;68:1-10.

22. Cremer K, Mackett M, Wohlenberg C, Notkins AL, Moss B. Vaccinia virus recombinants expressing herpes simplex virus type 1 glycoprotein D prevents latent herpes in mice. Science 1985;228;737-40.

23. Dea S., Bilodeau R, Sauvageau R, Montpetit C, Martineau GP. Antigenic variant of swine influenza virus causing proliferative and necrotizing pneumonia in pigs. J Vet Diagn Invest 1992;4:380-92.

24 Done SH, Brown IH. Swine influenza in the United Kingdom, past and present. Large Animal Practice 1997;Mar/Apr:20-8.

25. Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. Smallpox and its eradication. Geneva: World Health Organization; 1988.

26. Finney DJ. Spearman-Karfaer and moving averages. Statistical method in biological assay. 3rd ed. London: Charles Griffm & Company Ltd; 1978. p 394-401.

27. Gillard S, Spehner D, Drillien R. Mapping of a vaccinia virus host range sequence by insertion into the viral thymidine kinase gene. J Virol 1985;53:316-8.

28 Hinshaw VS, Alexander DJ, Aymard M and others. Antigenic comparisons of swine- influenza-like HlNl isolates from pigs, birds and humans: an international collaborative study. Bull World Health Organ 1984;62:871-8.

29 Hinshaw VS, Bean Jr. WJ, Webster RG, Easterday BC. The prevalence of influenza viruses in swine and the antigenic and genetic relatedness of influenza viruses from man and swine. Virology 1978;84:51-62.

30 Ito T, Nelson J, Couceiro SS, Kelm S, Baum LG, Krauss S, Castrucci MR, Donatelli I, Kida H, Paulson JC and others. Molecular basis for the generation in pigs of influenza A viruses with pandemic potential. J Virol 1998 Sep;72(9):7367-73.

31 Jones L, Ristow S, Yilma T, Moss B. Accidental human vaccination with vaccinia virus expressing nucleoprotein gene. Nature 1986 Feb;319(6054):543.

32 Kaplan MM, Webster RG. The epidemiology of influenza. Sci Am 1977;237:88-106.

33 Karber G. Beitrag zur kollektiven behandlung pharmakologischer reihenversuche. Arch Exp Pathol Pharmakol 1931;162:480-3. 99

34. Kida H, Ito T, Yasuda J, Shimizu Y, Itakura C, Shortridge KF, Kawaoka Y, Webster RG. Potential for transmission of avian influenza viruses to pigs. J Gen Virol 1994;75:2183-8.

35 Kundin WD. Hong Kong A2 influenza virus infection among swine during a human epidemic in taiwan. Nature 1970;228:857.

36 McMichael AJ, Gotch FM, Noble GR, Beare PAS. Cytotoxic T-cell immunity to influenza. N Engl J Med 1983;309:13.

37, Meyer H, Sutter G, Mayr A. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J Gen Virol 1991;72:1031-8.

38. Moss B. Vaccinia virus: a tool for research and vaccine development. Science 1991 Jun;252:1662-7.

39. Murphy BR, Webster RG. Orthomyxoviruses. Fields BN, Knipe DM, Howley PM and others, eds. Fields Virology. 3rd ed. Philadelphia: Lippincott-Raven Publishers; 1996. p 1397-445.

40. Oguiura N, Spehner D, Drillien R Detection of a protein encoded by the vaccinia virus C7L open reading frame and study of its effect on virus multiplication in different cell lines. J Gen Virol 1993;74:1409-13.

41. Ohishi K, Maruyama T, Shida H, Nishimaki J, Miki K, Sagata N, Ikawa Y, Sugimoto M. Immunogenicity of a recombinant vaccinia virus expressing envelope a glycoprotein of bovine leukaemia virus. Vaccine 1988 Oct;6:428-32.

42. Olsen CW, McGregor MW, Cooley AJ, Schantz B, Hotze B, Hinshaw VS. Antigenic and genetic analysis of a recently isolated HlNl sAvine influenza virus. Am J Vet Res 1993 0ct;54(10):1630-6.

43. Panicali D, Davis SW, Weinberg RL, Paoletti E. Construction of live vaccines by using genetically engineered poxviruses: biological activity of recombinant vaccinia virus expressing influenza virus hemagglutinin. Proc Natl Acad Sci U S A 1983;80:5364-8.

44. Paoletti E, Lipinskas BR, Samsonoff C, Mercer S, Panicali D. Construction of live vaccines using geneticallly engineered poxviruses: biological activity of vaccinia virus recombinants expressing the hepatitis B virus surface antigen and the herpes simplex virus glycoprotein D. Proc Natl Acad Sci USA 1984;81:193-7. 100

45 Perkus ME, Goebel SJ, Davis SW, Johnson GP, Limbach K, Norton EE, Paoletti E. Vaccinia virus host range genes. Virology 1990;179:276-86.

46 Perkus ME, Piccini A, Lipinskas BR, Paoletti E. Recombinant vaccinia virus: immunization against multiple pathogens. Science 1985;229:981-4.

47 Reed LJ, Muench H. A simple method of estimating 50% endpoints. American Journal of Hygiene 1938;27:493-7.

48 Rekik MR, Arora DJS, Dea S. Genetic variation in swine influenza virus A isolate associated with proliferative and necrotizing pneumonia in pigs. J Clin Microbiol 1994 Feb;32(2):515-8.

49 Schmidt NJ, Emmons RW. General principles of laboratory diagnostic methods for viral, rickettsial, and chlamydial infections. Schmidt NJ, Emmons RW, eds. Diagnostic procedures for viral, rickettsial, and chlamydial infections. 6th ed. Washington, D. C.: American Public Health Association; 1989. p 1-35.

50, Scholtissek C, Burger H, Bachmann PA, Harmoun C. Genetic relatedness of hemagglutinins of the HI subtype of influenza A viruses isolated from swine and birds. Virology 1983;129:521-3.

51. Schultz U, Fitch WM, Ludwig S, Mandler J, Scholtissek C. Evolution of pig influenza viruses. Virology 1991;183:61-73.

52. Shortridge KF, Webster RG, Butterfield WK, Campbell CH. Persistence of Hong Kong influenza virus variants in pigs. Science 1977;196:1454-5.

53. Smith GL, Bermink JR, Yewdell JW, Small PAJr., Murphy BR, Moss B. Vaccinia virus recombinants expressing influenza virus genes. Options for the control of influenza. Alan R. Liss, Inc.; 1986.

54. Suarez DL, Perdue ML, Cox N, Rowe T, Bender C, Huang J, Swayne DE. Comparisons of highly virulent H5N1 Influenza A viruses isolated from humans and chickens from Hong Kong. J Virol 1998 Aug;72(8);6678-88.

55. Sutter G, Moss B. Nonreplicating vaccinia vector efficientiy expresses recombinant genes. Proc Natl Acad Sci USA 1992 Nov;89:10847-51.

56. Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 1994;12(11): 1032-40. 101

57. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG. Initial genetic characterization of the 1918 "Spanish" influenza virus. Science 1997 Mar,275:1793-6.

58. USDA. Veterinary Biological Products. Ames, lA: USDA/APHIS Center for Veterinary Biologies; 1998.

59. Wraith CD, Vessey AE, Askonas BA. Purified influenza virus nucleoprotein protects mice from lethal infection. J Gen Virol 1987;68:433-40.

60. Wyatt LS, Shors ST, Murphy BR, Moss B. Development of a replication-deficient recombinant vaccinia virus vaccine effective against parainfluenza virus 3 infection in an animal model. Vaccine 1996;14(15):1451-8.

61. Yuen KY, Chan PKS, Peiris M, Tsang DNC, Que TL, Shortridge KF, Cheung PT, To WK, Ho ETF, Sung R and others. Clinical features and rapid viral diagnosis of human disease associated with avian influenza A H5N1 virus. Lancet 1998 Feb;3 51:467-71. 102

-ss RNA rt j

-SS RNA cDNA

PGR

NP HI H2

CloneAmp® +HA or NP, OSequenced i JS5 • HA. NP * I NP JS5+NP Then +HA

G06 • NP-P-P-HA * i NPjiHA

* screen by PGR, confirm by REA

Figure I - Diagrammatic representation of the procedure used to insert flA and NP into plasmid vector G06 (-ss = negative sense single strand RNA; RT = reverse transcription; CloneAmp, JS5 = plasmid vectors, PP = double promoter in opposite orientation). 103

BamHI A Pstl

EcoRI

5.3 Kb EcoR1

BaraHl \ V ^ NP Pstl

'.J\sc1 ..Pstl -EcoRI ..BamHI ...SnaB1-Not1...*

B G06-NP-HA RE Analysis

BamHI Pst1 EcoR1

1 .B 2.5 0.9 6.S 5.4 0.9 0.3 0.9 1.5 0.2 1.4- 4.1

Figure 2 - A) G06 is a pUC-based vector containing flanking MVA sequences and double promoter (PP). With the insertion of NP and HA, restriction enzyme (RE) sites provide confirmation of proper orientation. B) Digestion of plasmid DNA with Smal, Bam HI, Pstl, and EcoRl results in the fi^agmentation pattern desired. 104

Figure 2 - Continued; C) Electrophoresis following RE digestion discloses the expected pattern. 105

NP HA

G06-NP-HA (4 Clones)

MVA-lnfected CEFs

pass, IPA gpt • plaque purification > 6x

Figure 3 - Representation of the procedure used for transfection of MVA-infected cells, passage on cells, and subsequent plaque purification for at least 6 rounds (IPA = immunoplaque assay, gpt = gpt selection media). 106

Figure 4 - A) Photomicrograph of a fixed and stained MVA/SIV plaque in an early stage of viral expansion. Primary antibody used was hyperimmune porcine-origin SIV antisera. B) A live immunostained MVA/SIV plaque (not fixed), also using hyperimmune porcine-origin SrV antisera as primary antibody. C) A fixed and stained MVA/SIV plaque, using HA- specific monoclonal antibody. D) A fixed and stained MVA/SIV plaque, using NP-specific monoclonal antibody. Scale bar =100 microns. 107

Table 1 - A) Scoring of clinical signs during days 1-5 postchallenge. B) Results of scoring. A Clinical Signs to be Noted Points Given per Day Nasal and Ocular Discharge Serous 1 Slight Mucopurulent 2 Moderate " J Heavy 4 Depression, Coughing, Sneezing, and Abnormal Breathing Mild I Moderate 2 Severe J-» Gauntness, Dehydration 1 Temperature > 104.5T 1

Cumulative Clinical Sign Score for Pigs in Different Groups SIGN CHALLC PR8 IM SIVIM PR8 IN SIV IN Nasal discharge 2 Depression 30 2 Coughing J 2 J Sneezing 11 I 1 4 Abnormal Breathing 16 J-» Fever>104.5T 7 2 TOTAL 69 8 1 9 0

CHALL C = challenged nonvaccinated pigs; PR8 IM = MVA/PR8 IM vaccinates; SIV IM = MVA/SIV IM vaccinates; PR8 EST = MVA/PR8 IN vaccinates; SIV IN = MVA/SIV IN vaccinates. 108

Table 2 - Summary of results for lung lesions 5 days postchallenge in vaccinated and nonvaccinated pigs.

Pig Groups Variable CHALL C PR8 IM srviM PR8 IN SIV IN

# of pigs with gross lung lesions 6/6 6/6 0/6 6/6 2/6

% of lung < 10 <10 involvement 2-30 1 with 15 0 I with 15 minimal

presence of virus 3/6 2/6 in the lung (by EHC) 6/6 small 0/6 0/6 small amount amount

histopathological lung lesions 6/6 6/6 1/6 6/6 3/6

CHALL C = challenged nonvaccinated pigs; PR8 IM = MVA/PR8 EM vaccinates; SIV IM = MVA/SIV IM vaccinates; PR8 IN = MVA/PR8 IN vaccinates; SIV IN = MVA/SIV IN vaccinates.

HC = immunohistochemistry. 109

Table 3 - Summary of results for viral isolation on days 2-5, nasal shedding on days 4 and 5, and titer of virus shed on day 4 postchallenge in vaccinated and nonvaccinated pigs.

Pig Groups Variable CHALLC PR8 IM srviM PR8 IN SIV IN

Cumulative virus Isolations Days 2-5 23/24 20/24 11/24 15/24 7/24 6 pigs/group I swab/day/pig

# Shedding IN 4 DPC 6/6 6/6 1/6 6/6 3/6

# Shedding IN 5 DPC 5/6 2/6 1/6 0/6 0/6

Average (logio) IN titer (EIDso/ml) J 2 0.8 1.5 0.5 4 DPC

CHALL C = challenged nonvaccinated pigs; PR8 IM = MVA/'PRS EM vaccinates; SIV IM = MVA/SrV IM vaccinates; PR8 IN = MVA/PR8 EST vaccinates; SIV IN = MVA/SIV IN vaccinates.

IN = intranasal; IM = intramuscular; DPC = days postchallenge. 110

Table 4 - Immune response in pigs as measured by geometric mean titer (GMT) of HAI titers; pig sera vs. PR8 and ISU influenza virus antigen

Antibody Titer at Different Times Day 0 PVl 14 DPVl -1 DAY PC 5 DAYS PC c. 21 do Day 0 PV2 c. 14 DPV2 c. 56 do PIG #OF PR8 ISU PR8 ISU PR8 ISU PR8 ISU GROUP PIGS

CHALL C 6 ------

MVA/PR8IM 6 - - 22 - 508* - 452* -

MVA/SIVIM 6 - - - - - 57 - 113

MVA/PR8IN 6 - - 2 - 27 - 34 -

MVA/SrVIN 6 - - - - - 40 - 160

NEGC 2 - - - - nd nd - -

* On -1 and 5 days postchallenge, titers of pigs vaccinated with MVA/PR8 IM are significantly greater than those for the MVA/PR8 IN group. There is no significant difference in titers between the two MVA/SIV groups for those days.

PVl = postvaccination 1; PC = postchallenge; PV2 = postvaccination 2; do = days old, CHALL C = challenged controls, NEG C = nonchallenged controls. Ill

C3IS/IT eoo

soo

300

1 oo

o o 5 1 o -I s 30 35 DAY

-^PR8-IM/PR8 -f-SIV-IM/ISU -^PR8-IN/PR8 -°-SIV-IN/ISU

Figure 5 - Immune response in pigs as measured by geometric mean titer (GMT) of group hemagglutination mhibition (HAI) titers. This graph represents data found in table 4. Following second vaccination at day 14, the HAI titer rose markedly only for the group of pigs receiving the human flu recombinant IM. Following SIV challenge, there was an increase in titer for the pigs inoculated with MVA/SIV, but the titers for the MVA^RS IM group were not enhanced. 112

Table 5 - Immune responses in pigs as measured by geometric mean titer (GMT) of serum neutralization titers: pig sera vs. MVA and homologous SFV

Antibody Titer at Different Times Day 0 PVI 14 DPVl -1 DAY PC 5 DAYS PC c. 21 do Day 0 PV2 c. 14 DPV2 c. 56 do PIG #OF MVA SIV MVA SIV MVA SIV MVA SIV GROUP PIGS

CHALL C 6 ------

MVA>^R8 [M 6 - - - - 1.4 - 1.3 4.5

MVA/SrVIM 6 - - - - - 2.1 1.3 7.0

MVA/PR8IN 6 ------5.2

MVA/SIV IN 6 - - - - - 1.3 - 3.9

NEGC 2 ------

On Day 5 post challenge, each of the vaccinated pig groups had significantly higher GMT (p<0.05) than the nonvaccinated controls, but there were no significant differences between the vaccinated groups.

PVl = postvaccination 1; PC = postchallenge; PV2 = postvaccination 2; do = days old, CHALL C = challenged controls, NEG C = nonchallenged controls. 113

3. EVALUATION OF INHIBITION OF IMMUNE RESPONSE TO VACCINATION FOLLOWING SEQUENTIAL USE IN PIGS OF TWO MODIFIED VACCINIA VIRUS ANKARA (MVA) RECOMBINANTS EXPRESSING HETEROLOGOUS INFLUENZA VIRUS GENES

A paper to be submitted to Veterinary Microbiology

P. L. Foley, 1. R. Stoll, S. K. Hanson, L. A. Wilbur, P. S. Paul

Abstract

The objective of this study was to detennine whether there is an inhibition in immune response to secondary vaccination, following sequential immunization with two vaccinia virus recombinants. The recombinants were constructed from the highly attenuated modified vaccinia virus Ankara strain (MVA), and contained the hemagglutinin (HA) and nucleoprotein (NP) genes of either a human strain or a porcine strain of influenza virus.

Sixteen colostrum-deprived, caesarean-derived pigs were divided into three groups. One group of six was immunized with the recombinant human influenza virus vaccine intramuscularly (IM) and another group of six received it intranasally (IN) at both 21 and 35 days of age. Four pigs served as unvaccinated controls. At 91 and 105 days of age, all 16 pigs were administered EM the recombinant vaccine virus containing the swine influenza virus

(SrV) HA and NP. Blood samples were collected at 2-week: intervals through 23 weeks of age, and tested for antibodies by hemagglutination inhibition (HAI) against both strains of influenza virus and by serum neutralization (SN) with MVA and SIV. There was no significant inhibition of neutralizing antibodies during the study. Secondary SIV HAI responses were significantly greater following initial IN vaccination. Serum titers that developed against the human influenza virus, PR8, did not prevent development of SIV titers. 114

This suggests that sequential use of such recombinants containing inserts from various strains may be successful in practice.

1. Introduction

The advent of recombinant vaccines capable of expressing numerous foreign inserts has raised the question of whether frequent use of such constructs to deliver different antigens might be inherently nonproductive (Etlinger & Altenburger 1991; Flexner, et al. 1988;

Rooney, et al. 1988). Would, in fact, preexisting immunity to the virus adversely affect efficacy? It has been thought by some that the immune response generated against a vector would limit its replication and prevent adequate expression of the inserts. This, in turn, would inhibit the induction of protective immunity against the heterologous proteins. In support of this view, one study (Chelyapov, et al. 1988) demonstrated the strength of the immune response to W. It showed that antibodies were generated against most of the vaccinia virus (W) structural proteins, including those proteins located internally within the virus, in both rabbits and humans, that these antibodies in humans were preserved over many years, and that insertion of foreign genes did not affect this pattern. Additional reports indicated that preexisting immunity to W resulted in reduced titers of antibody to the foreign protein (Cooney, et al. 1991; Rooney, et al. 1988), decreased protection (Rooney, et al.

1988), and reduced, transient T-cell response (Cooney, et al. 1991). Another study (Kundig, et al. 1993) demonstrated that vaccination with one W recombinant resulted in long term suppression of the humoral response in mice to a second W recombinant's gene product. 115

On the other hand, there have been reports of boosting antibody titers in animals given the same recombinant virus, or of induction of immunity against a foreign antigen in animals vaccinated with W subsequently inoculated v^th a recombinant W expressing that antigen

(Etlinger & Altenburger 1991; Jones, et al. 1986; Perkxis, et al. 1985; Rooney, et al. 1988).

Also, antibody titer to one of the heterologous proteins in a multivalent W recombinant did not prevent mice from developing immunity (Rexner, et al. 1988). Still, questions have persisted about the adequacy of the response in these cases, even if outright suppression has not occurred.

To avoid interference to secondary vaccination and improve the immune response, some have tried using diversified prime and boost regimens for vaccinations, as in priming with a

W recombinant, followed by a subunit vaccine containing the expressed protein. Use of a subunit as booster has met with mixed results (Cooney, et al. 1993; VIontefiori, et al. 1992).

In one study, use of a W recombinant, boosted by a recombinant avian pox virus, resulted in improved T-cell responses, better than when using either construct alone (Hodge, et al.

1997). In another study, mice primed with an influenza virus recombinant and boosted with a

W construct expressing the same antigen developed strong secondary antigen-specific CD8'

T cell responses, but only if the vaccines were given in just that order. However, these regimens are not appropriate for achieving widespread prophylaxis and therapy on a commercial scale. From a practical standpoint, the preferred alternative would be to have just one vector, capable of being used safely and repeatedly to express numerous inserts.

In the present study, the concept of repeated use of the same W vector containing different inserts was explored. Unlike other studies, however, the vector chosen was the 116 modified vaccinia virus Ankara (MVA) strain, which was safely used in over 120,000 people during the smallpox eradication program, and which has been well characterized (Bender, et al. 1996; Carroll & Moss 1997; Carroll, etal. 1997; Meyer, et al. 1991; Sutter & Moss 1992;

Sutter, et al. 1994; Wyatt, et al. 1996). MVA has been used successfully in the past as a vector for influenza virus structural proteins in mice (Bender, et al. 1996; Sutter, et al. 1994), parainfluenza virus 3 proteins in cotton rats (Wyatt, et al. 1996), and simian immunodeficiency virus env-gag-pol in macaques (Hirsch, et al. 1996). It was also able to protect against and provide therapy for pulmonary metastases in mice (CarrolL et al. 1997).

The MVA strain is replication-incompetent in most all mammalian cells tested, a consequence of six major deletions equal to 15% of its genome that prevents virus assembly but leaves gene expression, both early and late, relatively unimpaired. This feature could presumably allow repeated use of the vector with little induction of inhibiting levels of anti-vaccinia virus antibody. If a protective immune response could be generated against an inserted protein, even following repeated use of the MVA vector, then the value of MVA as a delivery system for immunogens would be increased.

To conduct this study, two MVA recombinants were used. The first of these contains the

HA and NP genes fi'om the HlNl human influenza strain, A/PR/8/34 (PR8), in opposite orientation and under the control of two optimized synthetic early/late promoters. This construct has been shown to provide protection to mice against a lethal challenge from PR8

(Sutter, et al. 1994). The second MVA-vectored recombinant, using the same promoters, expresses the HA and NP genes from a type A HlNl field strain of swine influenza virus

(SIV). These gene inserts were chosen because previous studies have demonstrated that HA 117

induces virus-neutralizing antibody, important for protection against infection (Webster, et al.

1991); and that NP is the major target antigen recognized by type-specific cross-reactive

cytotoxic T-lymphocytes (CTL) (Townsend & Skehei 1984; YewdeU, etai 1985),

responsible for recovery firom infection (Wraith, et al. 1987).

Since most of the protective humoral response following exposure to influenza virus is generated against the five hypervariable antigenic sites of HA, antibodies to this protein

provide immunity to influenza viruses with a sufficiently similar HA (Andrew, et al. 1987).

In a previous study at our laboratory, it was determined that the PR8 HA is not homologous enough to the SIV HA for their respective antibodies to be cross-reactive on serological assays (Foley et al, manuscript in preparation). This lack of reactivity allows, in the present study, an opportunity to evaluate MVA antibody interference, using identical test methodology to determine antibody response against heterologous gene inserts.

The optimal route of vaccine delivery for a respiratory pathogen, such as influenza, has been a matter of some discussion. One study (Small, et al. 1985) found that intranasal (IN) administration of a W recombinant expressing HAl in mice protected both lung and nose against homologous challenge, but scarification protected only limg. Another group (Meitin, et al. 1991) vaccinated mice either intraperitoneally (IP) with a killed PR8 vaccine, stimulating high serum IgG, or IN with a W recombinant containing HAi, inducing nasal

IgA titers. They found that the lungs but not noses in the IP group and the noses but not lungs in the IN group were fiiUy protected against challenge. By reversing protection in mice

recovered fi-om infection with influenza, using anti-IgA antiserum, it has since been shown 118 that IgA is the primary factor in nasal immunity to influenza virus in mice (Renegar & Small

Jr. 1991). To maximize both IgA. and IgG, several studies have tried to stimulate the

'common mucosal immune system' (Czerkinsky & Svermerholm 1993; Yetter, etal. 1980) in mice by intragastric or intrajejeunal administration of a W or MVA recombinant expressing influenza virus genes (Bender, et al. 1996; Meitin, et al. 1994). They, in fact, have been successful in attaining mucosal IgA, serum IgG, and CTL activity. However, the use of orally-administered enteric-coated capsules is not a viable option for weanling pigs, at this time, though they hold great promise for human vaccination programs. Consequently, the pigs used in this study were inoculated either intramuscularly (IM) or IN, to evaluate the effect these routes might have on inhibition of secondary vaccination.

2. Materials and methods

2.1 Cells and Viruses including MVA recombinants

Vaccinia virus strains MVA and the MVA recombinant expressing PR8 HA and NP genes

(MVA/PR8) were generously provided by B. Moss, Laboratory for Viral Diseases, National

Institute of Allergy and Infectious Diseases (NIAED), National Institutes of Health (NIH).

The MVA recombinant expressing SIV HA and NP genes (MVA/SFV) was recently constructed in our laboratory (Foley et al., manuscript in preparation) from an Iowa field strain of SIV isolated by the Iowa State University Veterinary Diagnostic Laboratory

(courtesy of B. Janke). MVA, MVA/PR8, and MVA/SIV were grown at 37°C and 5% CO, 119 in the second passage of chicken embryo fibroblast cells, using M199 medium with F-10 nutrient mixture (M199/F-10), supplemented with 0.15% bactotryptose phosphate broth,

0.09% Na bicarbonate, 1% L-glutamine, 25 U/mi penicillin G potassium, 75 U/ml streptomycin sulfate, 0.1% gentocin, and 5% FBS.

A 1960's SrV field isolate of the 'classical' swine HlNl subtj^je, provided by the

Diagnostic Virology Laboratory (DVL), National Veterinary Services Laboratories (NVSL); a reference 'classical' HlNl virus, A/Sw/IN/1726/88, provided by Virginia Hinshaw,

University of Wisconsin; and the ISU field isolate were propagated allantoically in 10-day-oId embryonated chicken eggs at 37°C for 72 hours, and titrated by mean egg infectious dose

(EID30) and hemagglutination (HA) units. In addition, several cell lines, including Madin-

Darby Canine Kidney (MDCK), swine testicle (ST), porcine kidney (PK-15), and another dog kidney line (DK-F), were tested for their ability to support SIV growth, with and without the use of centrifiigation and trypsin, using the three strains of SIV. It was determined that a

24-hr-old ST monolayer, seeded at 2 x 10^ cells/ml, and grown at 37°C, in 5% CO,, using

Minimum Essential Medium (MEM) with Earle's salts (Gibco cat. no. 41500-018), supplemented with 0.22% Na bicarbonate, 0.5% edamine, 1% L-glutamine, 25 U/ml penicillin G potassium, 75 U/ml streptomycin sulfate, 0.1% gentocin, 1% sodium pyruvate, and 5% FBS, provided good conditions for SIV replication (data not shown). At 24 hours afl;er seeding, medium was decanted and the ST monolayer inoculated with virus in sufficient medium to cover the monolayer. Following virus adsorption for one hour at 37°C, fi^esh medium was added and the cultures incubated at 37°C in 5% CO,. 120

2.2 Titration of viruses

For MVA and MVA recombinant virus titrations, 24-hour old CEF cells, seeded at 8 x 10^ cells/ml in 60 millimeter (mm) tissue culture plates under conditions cited above, were decanted then refed with new media containing only 1% FBS, inoculated with 100 ul/well of serial 10-fold dilutions of virus, incubated at 37°C in 5% CO,, then fixed at 2 days postinfection. MVA recombinant-inoculated plates were fixed in 1:1 acetone:methanol, reacted with primary antibody, then anti-mouse or anti-swine peroxidase-labeled conjugate.

The substrate used for these assays was metal enhanced DAB (Pierce, Rockford, EL).

Recombinant virus titers were determined using insert-reactive anti-HlNl SIV polyclonal antisera and anti-PR8 HA and NP monoclonal antibodies (courtesy of J. Yewdell, NIAID,

NIH) as primary antibody, and expressed as plaque-forming units (PFU). MVA titers were determined using either cytopathic effect (CPE) or immunofluorescent assay (IFA.). For the latter procedure, the plates were fixed in 80% acetone. The primary antibody was a rabbit- origin anti-W polyclonal antisera, reacted vsnth an anti-rabbit FITC-conjugate, with titers expressed as fluorescent focal units (FFU). This EFA procedure was also used on the MVA recombinant-infected plates, subsequent to the DAB immunoplaque assay, to determine the titer of non-expressing plaques.

For SIV titration on the ST cell line, 96-well plates were inoculated with 50 ul virus dilution per well, centrifiiged at room temperature for 2 hours at 400xg, incubated at 37°C in

5% CO2, then read by CPE or EFA, using anti-HlNl SIV polyclonal antisera, 5-7 days postinoculation. 121

2.3 Animal Immunizations

16 colostmm-deprived, caesarean-derived pigs (CDCD) were divided into 3 groups. One group of 6 received 1 mi of c. lO^PFXJ of the MVA/PR8 recombinant IM along with an additional c. 1.7 logio FFU of MVA not expressing the insert; another group of six received the same inoculum EN in 2 ml (diluted 1:1 in 0.01 M PBS, pH 7.2, 1 ml per nostril) at both 21 and 35 days of age. The IN inoculations were administered using a gas-powered atomizer, as previously described (Sinclair & Tamoglia 1972). Four pigs remained unvaccinated until 91 and 105 days of age, when all 16 pigs were administered 1 ml of c. 10^ PFU of the MVA/SIV recombinant IM, along with c. 0.1 logio FFU of non-insert-expressing MVA.

2.4 Serological assays

Serum samples were harvested at 2-week intervals beginning at 21 days of age, through

161 days of age, then analyzed by hemagglutination inhibition (HAI) against the two strains of influenza virus, PR8 and ISU, and by serum neutralization (SN) of MVA and SIV, using the constant virus-varying serum method.

HA and HAI tests were performed in 96-well microtiter plates, using 0.5% chicken erythrocytes. For the SIV and PR8 HAI assay, sera were pretreated using 10% kaolin and

The MVA SN procedure used 96-weIl plates containing 24-hour-old monolayers of CEF cells, inoculated with 50 ul of a 1:1 mixture of virus and 5-fold dilutions of serum, previously incubated at 37°C for 1 hour. Back titration established virus titer at 100-200 mean tissue culture infectious dose (TCID50). The inoculum was decanted after 1 hour, replaced with

200 ul of M199/F-10 media, supplemented as above but with only 0.5% FBS, and read at 5-7 days by CPE. SN titers were determined by the method of Reed-Muench (Reed & Muench

1938; Schmidt & Emmons 1989).

The SIV SN procedure, except for the use of the ST cell line, a different maintenance medium (MEM and Earle's salts with no FBS), and centrifugation as described above, is otherwise the same as that used for the MVA SN assay. The DVL strain of SFV, because better adapted to cell culture, was utilized in SN assays at 70-250 TCID50.

2.4 Statistical analysis

For comparison of data between groups within a given table and between MVA/PR8 EM, day 0 to 70 on Table la, and MVA/SIVIM, day 70 to 140 on Table lb (HAI titers), an analysis of variance (ANOVA) was conducted. The means were compared using the Least

Significant Difference (LSD) method. For all correlation statistics, the correlation coeflBcient was based on the logarithms of the HEAI and SN titers. To determine geometric mean titer

(GMT), logio of each titer was summated, then averaged and the antilog determined.

Because negative titers are '0', 1 was added to all numbers, then subtracted later firom the averaged GMT. 123

2.5 Sequence analysis

The ISU SrV HA and NP genes have been previously cloned and described (Foley, et aL, manuscript in preparation), but their nucleotide sequences have not been previously reported.

The cloned genes were subjected to primer walking along both strands of DNA, starting with the Universal (-21M13) and Reverse CR-2', M13-USB) primers, using the ABI Prism Model

377 DNA Sequencer (Perkin Ebner). Sequences generated were aligned, edited, and assembled using AutoAssembler (Perkin Elmer). HA and NP sequences from PR8 and

A/Sw/IN/I726/88 (IN88), a 'classical' t5T)e A HlNl SIV, were retrieved from the GenBank database using the respective accession numbers for PR8 HA (J02143) and NP (J02147), and

IN88 HA (M81707) and NP (L46849). Sequences were compared using Omiga 1.1 software

(Oxford Molecular Group). Slight alterations, consisting of a single base addition, were made in the retrieved sequences only when needed to preserve the reading frame. In one case, addition of a single 'T' at IN88 HA base 1629, converted an otherwise meaningless amino acid code into one that closely followed the code for ISU HA for the remainder of the protein sequence. Likewise, an apparent missing base at nucleotide 1630 in PR8 HA rendered the bases thereafter totally heterologous. If a 'missing' G (found in ISU and INS 8

HA) is added, the homology is 83.4% up to amino acid 556, when the code again becomes scrambled. If the first 1629 bases (or 543 amino acids) only are compared, that is, without adding the G, then the amino acid homology is 83.1%. It was felt that this was close enough to the first figure to call at 83% homology. 124

3. Results

3.1 Serological responses

Examination of tiie HAI titers against the PR8 strain over the 140-day study indicates no cross reactivity with the ISU strain, in that PR8 titers were not enhanced after inoculation with MVA/SrV on days 70 and 84 (Table 1, Figure 1). In fact, HAI titers among the

MVA/PR8 IM group continued to decline. Also, the MVA/PR8 IM titers are considerably higher than in the IN group after the initial vaccinations.

Following subsequent IM vaccination in all 16 pigs with MVA'SIV, HAI titers against ISU rose faster and higher (p<0.05) in the pigs previously vaccinated EN with the MVA/PR8 recombinant (Table2, Figure 2). The increase is similar to that seen in the MVA/PR8 IM group following vaccination with MVA/PR8 by IM route on days 0 and 14 (Table 3).

There is no detectable SN titer against MVA in the MVA^R8 IN group until two weeks after administration of the first MVA/SIV inoculation (Table 4, Figure 3). Even then, those

MVA SN titers are less than in pigs having received MVA'PRS by IM route (p<0.05). The

MVA SN titers in the MVA/PR8 IM group appear by two weeks after the second MVA'PRS vaccination, and are boosted by the MVA/SIV injections on days 70 and 84.

The SN titers against SFV are not apparent until two weeks after the first MVA/SIV vaccination (Table 5, Figure 4). At that time there is no significant difterence in SN between 125 pigs previously vaccinated either IM or IN with the MVA/PR8 construct, but both groups have a significantly greater response than those in the control group vaccinated only with

MVA/SIV (p<0.05).

3.2 Statistical analysis

The PR8 HAI titers of the MVA/PR8 IM group fi-om day 0 to day 70 (Table 1) show a similar rise in HAI titer and, in fact, are not statistically different using ANOVA/LSD methodology, fi-om the ISU HAI titers of the MVA/PRS group from day 70 to day 140, following MVA/SrV IM vaccination (Table 2). A qualitative representation, comparing the relative immune responses of the groups, is depicted in Table 3. Also, there is no negative correlation between MVA SN titers (Table 4) and ISU HAI titers (Table 2) for days 84 to

140, that is, a higher MVA SN titer is not associated with a lower ISU HAI titer. Regarding the SIV SN titers (Table 5), both the IM and EN groups were significantly (p<0.05) greater than the control group but not different fi-om each other.

3.3 Sequence analysis

The ISU HA gene had 98% homology to the IN88 HA at the nucleotide level, but only

81% to the PR8 HA (Table 6). This represented 99% homology at the amino acid level between ISU and IN88 HA, but only 83% to the PR8 HA. Regarding comparison of NP, 126

ISU and INSS shared 96% nucleotide and 99% amino acid similarity; the PR8 NP, when compared to that of ISU, showed 87% nucleotide and 92% amino acid homology. The differences between the two genes of IN88 and ISU (figures 5a, 5b, 5c) did not appear to be at locations previously associated with changes in virulence, such as the cleavage site between the HI and H2 portions of the HA gene (Senne, et al. 1996).

4. Discussion

Both the MVA/PR8 IN and IM groups seem to have a rise in SIV SN titer following

MVA/SrV IM inoculation, not shared by the controls. It may be that a nonspecific immune stimulation due to previous vaccination occurred. This phenomenon, or some other, such as better priming of T helper cells, may be at play in the HAI titers against ISU, wherein the IN group titers appear to be increased when compared to those of the control and IM groups.

The IM group HAI titers are equivalent to those of the nonvaccinated controls and, thus, do not suggest interference. Yet, one might argue that the EM group titers are not as high as those of the IN group due to interference firom higher anti-MVA IgG titers, as suggested by the IM group's higher MVA SN titers. However, it cannot truly be said to be interference if the titers achieved are the same as those realized in naive animals. What's more, the ISU SN titers for the IM and IN groups are not significantly different, though statistically greater than the controls. The expectation would be that any interference would result in a decreased ability to neutralize virus. Instead, neither MVA nor SIV SN immune responses showed significant inhibition during the study (p<0.05). Lastly, and of great importance, the \n development of HAI titers against PR8 had no detectable effect on the development of HAI titers against the ISU strain.

Sequence analysis of the ISU and PR8 HA and NP genes indicates that, whereas the type

A-specific NP gene is fairly homologous (87% nucleotides, 92% amino acid residues), the more strain-specific HA gene is suflBciently heterologous (only 81% nucleotides are the same,

83% amino acids) to explain the lack of cross-reactivity on HAI assay. Sequential vaccination with such heterologous constructs should allow for independent titers, not inhibited by MVA titers. For more homologous constructs, the effect would seem to be, from the results seen here of several vaccinations, simple boosting.

Previous studies using replication-competent W strains have demonstrated interference caused by preexisting immunity to the W vector (Cooney, et al. 1991; Kundig, et al. 1993;

Rooney, et al. 1988). Presumably, the ability of these MVA recombinants to vector different inserts without secondary vaccination inhibition is due to MVA's host restriction. The supposition would be that, by not generating too strong a response, subsequent vectored vaccines are not overv^'helmed. This hypothesis should be tested further with different inserts and varied vaccination regimens.

One concern related to vaccination interference has been that, even where anti-W antibody has little effect on the antibody response to a foreign gene product, there may still be interference from pre-existing maternal antibody to the virus from which the protein derives. Some have noted that passively administered polyclonal antibody to a foreign protein can inhibit B-cell response to a W recombinant vaccine, but not necessarily the cytotoxic T-lymphocyte response (Galletti, etai 1995; Johnson, et al. 1993). One study 128

(Brockmeier, et al. 1997) found that, of two W recombinants expressing one of two pseudorabies glycoproteins, only one protected similarly regardless of the presence of maternal antibody, suggesting that choice of insert can be crucial to induction of immunity. It would be interesting to see if the replication-incompetent MVA can improve protection in the face of maternal passive immunity.

Researchers have reported success in culturing influenza viruses in various cell lines, including MDCK (Hinshaw, et al. 1994). In our hands, the ST cell line allowed for titration of SrV by both CPE and BFA, as did the DK-F line. The PK-15 cell line did not demonstrate apparent cytolytic effect upon infection with SIV, but virus replicated as evidenced by IF A.

In our laboratory, the MDCK line did not indicate the effects of SIV infection as readily or reliably when compared to the other cell lines.

References

Andrew ME, Coupar BEH, Boyle DB, Ada GL. 1987. The roles of influenza virus haemagglutinin and nucleoprotein as protective antigens against influenza virus infection in mice. Scand J Immunol 25:21-8.

Bender BS, Rowe CA, Taylor SF, Wyatt LS, Moss B, Small Jr. PA. 1996. Oral immunization with a replication-deficient recombinant vaccinia virus protects mice against influenza. J Virol 70(9):6418-24.

Brockmeier SL, Lager KM, Mengeling WL. 1997. Vaccination with recombinant vaccinia virus vaccines expressing glycoprotein genes of pseudorabies virus in the presence of maternal immunity. Vet Microbiol 58:93-103.

Carroll MW, Moss B. 1997. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia viras: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238:198-211. 129

Chelyapov NV, Antonova TP, YanovaNN, Chemos VI. 1988. Antigenic properties of vaccinia virus and of the virus recombinant strains expressing heterologous genes. Acta Virol (Praha) 32:409-16.

Cooney EL, McElrath MJ, Corey L, Hu SL, et al. 1993. Enhanced immunity to HIV envelope elicited by a combined vaccine regimen consisting of priming with a vaccinia recombinant expressing HIV envelope and boosting with gpl60 protein. Proc Natl Acad Sci U S A 90(5): 1882-6.

Czerkinsk>' C, Svennerholm A-M. 1993. Induction and assessment of immunity at enteromucosal surfaces in humans: implications for vaccine development. Clin Infect Dis 16(Suppl 2):S106-16.

Etlinger HM, Altenburger W. 1991. Overcoming inhibition of antibody responses to a malaria recombinant vaccinia virus caused by prior exposure to wild type virus. Vaccine 9:470-2.

Flexner C, Murphy BR, Rooney JF, Wohlenberg C, Yuferov V, Notkins AL, Moss B. 1988. Successful vaccination with a polyvalent live vector despite existing immunity- to an expressed antigen. Nature 335:259-62.

Galletti R, Beauverger P, Wild TF. 1995. Passively administered antibody suppresses the induction of measles virus antibodies by vaccinia-measles recombinant viruses. Vaccine 13(2):197-201.

Hinshaw VS, Olsen CW, Dybdahl-Sissoko N, Evans D. 1994. Apoptosis: a mechanism of cell killing by influenza A and B viruses. J Virol 6S(6):3667-73.

Hirsch VN, Fuerst TR, Sutter G, et al. 1996. Patterns of viral replication correllate with outcome in simian immunodeficiency virus-infected macaques: effects prior to immunization with a trivalent SIV vaccine in modified vaccinia virus Ankara (MVA). J Virol 70:3741-52. 130

Johnson MP, Meitin CA, Bender BS, Small Jr. PA. 1993. Recombinant vaccinia immunization in the presence of passively administered antibody. Vaccine 11(6): 665-9.

Jones L, Ristow S, Yilma T, Moss B. 1986. Accidental human vaccination with vaccinia virus expressing nucleoprotein gene. Nature 319(6054);543.

Kundig TM, Kalberer CP, HengartnerH, Zinkemagel RM. 1993. Vaccination with two different vaccinia recombinant viruses: long-term inhibition of secondary vaccination. Vaccine 11(11): 1154-8.

Meitin CA, Bender BS, Small Jr. PA. 1991. Influenza immunization: intranasal live vaccinia recombinant contrasted with parenteral inactivated vaccine. Vaccine 9: 751-6.

Meitin CA, Bender BS, Small Jr. PA. 1994. Enteric immunization of mice against influenza with recombinant vaccinia. Proc Natl Acad Sci U S A 91:11187-91.

Meyer H, Sutter G, Mayr A. 1991. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J Gen Virol 72: 1031-8.

Montefiori DC, Graham BS, Kliks S, Wright PF, etal. 1992. Serum antibodies to HIV-1 in recombinant vaccinia virus recipients boosted with purified recombinant gp 160. J Clin Immunol 12(6);429-39.

Perkus ME, Piccini A, Lipinskas BR, Paoletti E. 1985. Recombinant vaccinia virus: immunization against multiple pathogens. Science 229:981-4.

Reed LJ, Muench H. 1938. A simple method of estimating 50% endpoints. American Journal of Hygiene 27:493-7.

Renegar KB, Small Jr. PA. 1991. Immunoglobulin A mediation of murine nasal anti- influenza virus immunity. J Virol 65(4):2146-8.

Schmidt NJ, Emmons RW. 1989. General principles of laboratory diagnostic methods for viral, rickettsial, and chlamydial infections. Schmidt NJ, Emmons RW, eds. Diagnostic procedures for viral, rickettsial, and chlamydial infections. 6th ed. Washington, D. C.: American Public Health Association, p 1-35.

Senne DA, Panigrahy B, Kawaoka Y, Pearson JE, Suss J, Lipkind M, Kida EL, Webster RG. 1996. Survey of the hemagglutinin (HA) cleavage site sequence of H5 and H7 avian influenza viruses: amino acid sequence at the HA cleavage site as a marker of pathogenicity potential. Avian Dis 40:425-37.

Sinclair LR, Tamoglia TW. 1972. Intranasal administration of virulent viruses to cattle, with a gas-powered atomizer. Am J Vet Res 33(10):2085-6.

Sutter G, Moss B. 1992. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc Natl Acad Sci USA 89:10847-51.

Townsend ARN-I, Skehel JJ. 1984. The influenza A virus nucleoprotein gene controls the induction of both subtype specific and cross-reactive cytotoxic T cells. J Exp Med 160:552-63.

Webster RG, Kawaoka Y, Taylor J, Weinberg R, Paoletti E. 1991. Efficacy of nucleoprotein and haemagglutinin antigens expressed in fowlpox virus as vaccine for influenza in chickens. Vaccine 9:303-8.

Wraith CD, Vessey AE, Askonas BA. 1987. Purified influenza virus nucleoprotein protects mice from lethal infection. J Gen Virol 68:433-40.

Yetter RA, Lehrer S, Ramphal R, Small Jr. PA. 1980. Outcome of influenza infection:effect of site of initial infection and heterotj'pic immunity. Infect Immun 29(2):654-62. 132

Yewdell JW, Bennink JR, Smith GL, Moss B. 1985. Influenza A virus nucleoprotein is a major target antigen for cross-reactive anti-influenza A virus cytotoxic T lymphocytes. Proc Natl Acad Sci USA 82:1785-9. Table 1 Immune responses in pigs against PR8 strain, as measured by hemagglutination inhibition

Geometric mean HAI titers at days post first inoculation with IV1VA/PR8

Group n 0» 14'' 28 42 56 70*^^ 84" 98 112 127 140

IN ' 6 neg neg 0.5 0,5 0,5 2,7 1,2 1,2 2,7 2,3 0.5

IIM 6 neg 4,1 89,9 36,0 25,6 11.1 6.7 5,9 5,9 5,9 5,2

Control 4 neg neg neg neg neg neg neg neg neg neg neg

12 pigs were inoculated with MVA/PR8 recombinant, either IN or IM, Four controls did not receive the MVA/PR8 inoculum, all pigs, including the 4 controls not yet inoculated, were given the MVA/SIV recombinant by the IM route. Up through day 140, the last day of testing, the 4 controls had no detectable HI titer (< 10) against the PR8 strain of influenza virus, HAI: hemagglutination inhibition, n; number in group, neg: geometric mean titer = 0, 134

GMT

100-

0.1 ^ ^^^^^^ O 14 28 42 56 70 84 98 112 127 140 Day

~^IN ~+"(M "^Control

Figure 1. Immune response of pigs represented by GMT of group HAI titers against the PR8 strain of influenza virus. This graph corresponds to the data found in table 1. The response in the PR8 IM group was greatest, and was not stimulated by vaccination with MVA/SIV on days 70 and 84. Table 2 Immune responses in pigs against ISU strain, as measured by hemagglutination inhibition

Geometric mean HAI titers at days post first inoculation with MVA/PR8

Group n 0" 14" 28 42 56 84' 98 112 127 140

IN 6 neg neg neg neg neg neg 9.2 71.3 44,9 40,1 25,2

IM 6 neg neg neg neg neg neg neg 14,8 9,2 8,1 11,2

Control 4 neg neg neg neg neg neg neg 20,0 11.9 11,9 5,0

12 pigs were inoculated with MVA/PR8 recombinant, either IN or IM, Four controls did not receive the MVA/PR8 inoculum, °''' all pigs, including the 4 controls not yet inoculated, were given the MVA/SIV recombinant by the IM route, HAI: hemagglutination inhibition, n: number in group, neg: geometric mean titer = 0, 136

GMT

1 OO

1 O

^ ^ ^ >tC DvC 14 28 42 S6 TO 84 98 112 127 140 Day

^IN -f-IM ^Control

Figure 2. Immune response of pigs represented by GMT of group HAI titers against the ISU strain of influenza virus. This graph corresponds to the data found in table 2. The response was greatest in the group having received the MVA/PR8 vaccine IN. 137

Table 3 Qualitative representation of HAI response to vaccinations on days 0, 14 with MVA/PR8 and days 70, 84 with MVA/SIV

ROUTE OF 1ST RESPONSE TO 1ST RESPONSE TO 2ND VACCINE VACCINATION VACCINE (MVA/PR8) (MVA/SIVgiven IM)

IN + +++

DVI +++ -H-

NONE - ++

+ = weak, ++ = moderate, +++ = strong. HAI = hemagglutination inhibition. Table 4 Immune responses in pigs, as measured by ability to neutralize the vaccinia virus MVA

Geometric mean SN titers at days post first inoculation with MVA/PR8

Group n 0" 14" 28 42 56 70*^ 84" 98 112 127 140

IN 6 neg neg neg neg neg neg 5,2 2,0 2,7 3,2 3,7

IM 6 neg 0.3 8.6 1,4 1,6 0,3 28,3 17,2 28,1 22,8 15,1

Control 4 neg neg neg neg neg neg neg neg neg neg neg

12 pigs were inoculated with MVA/PR8 recombinant, either IN or IM, Four controls did not receive the MVA/PR8 inocuium, all pigs, including the 4 controls not yet inoculated, were given the MVA/SIV recombinant by the IM route. Up through day 140, the last day of testing, the 4 controls had no detectable titer (< 5) against MVA. SN: serum neutralization, n: number in group. neg: geometric mean titer = 0, 139

GMT

1OO

-I O

0.1 X jtc —9te jte O 28 56 1 27" 1 40 Day

IN -i-IM -^Control

Figure 3. Immune response of pigs represented by GMT of group SN titers against the MVA strain of vaccinia virus. This graph corresponds to the data found in table 4. The response in the previously vaccinated groups, as measured by virus neutralization, was increased following vaccination vi^ith the second recombinant. Table 5 Immune responses in pigs, as measured by ability to neutralize homologous SIV

Geometric mean SN titers at days post first inoculation with MVA/PR8

Group n 0" 14" 28 42 56 70' 84" 98 112 127 140

IN 6 neg neg neg neg neg neg 5,2 14,6 14,5 22,6 21,2

IM 6 neg neg neg neg neg neg 6,2 7,3 11,6 12,1 12,1

Control 4 neg neg neg neg neg neg neg 2,0 2,0 3.2 2,5

GMX

1 OO

1 O

14 28 42 84 98 112 127- 140 Day

IN ~^IM -^Control

Figure 4. Immune response of pigs represented by GMT of group SN titers against a homologous HlNl strain of SFV. This graph corresponds to the data found in table 5. The response in the vaccinates, as measured by virus neutralization, was significantly greater than in the controls. 142

Table 6 Nucleotide (NT) and amino acid (AA) homology of HA and NP genes of ISU SIV isolate with those of PR8 and IN88 isolates

% Homology With Influenza virus gene ISU SIV HA ISU SIV NP NT AA NT AA

PR8HA 81 83 - -

PR8 NP - - 87 92

IN88 HA 98 99 - -

IN88 HP - - 96 99 143

Sequence Differences SW/IN/1726/88 vs. ISUDL Amino Nucleic Acids Acids HI (1032 bp): 0 13 H2(692bp): 4 17 NP (1490 bp): 3 53

B HA Amino Acid Differences in S74 AAs m -H2 cleavage at #344-345 AA Residue IN/88 ISUDL 403 valine isoleucine 491 threonine lysine* 505 aspcu-agine lysine* 549 glycine serine * basic AAs

NP Amino Acid Differences in 1490 NTs or 496 AAs

AA Residue IN/88 ISUDL 316 valine isoleucine 450 asparagine serine 482 asparagine serine

Figure 5 - Sequence differences between the HA and NP genes of two SIV isolates. A) Although there is some diversity at the nucleotide level, the changes are nonessential. When translated, there are few amino acid (AA) differences between the two strains, the classic IN/88 and the ISU Diagnostic Laboratory (ISUDL) isolate. B) The ISUDL strain, from which the HA and NP genes were cloned, has a few more basic amino acids, but they are not at the cleavage site between HI and H2. An increase in the basic amino acids, lysine and arginine, at that location has been associated with increased virulence (Senne et al, 1996). C) The amino acid changes between the two NP proteins are few and apparently not significant. 144

GENERAL CONCLUSIONS

The first study undertaken in this thesis project demonstrated that the immunity provided by MVA/PR8 was incomplete against SIV infection, clinical disease, and virus shedding. The second study indicated the marked improvement in protection against homologous SrV challenge when pigs were vaccinated with the MVA/SFV construct. The third study provided evidence that sequential use of such MVA recombinants containing inserts fi-om various strains of influenza, given a mere two months apart, still generated appropriate immune responses to the different inserts. Taken together, the data underscore the suitability of the MVA vector to express immunogens in a safe, efficacious, and reproducible maimer. Although there is an inactivated SIV vaccine commercially available, there are numerous advantages to using a live recombinant vaccine. For one, the foreign protein expressed by the recombinant moves through the Class I pathway and is appropriately presented to cytotoxic T cells, priming cell-mediated immunity (Bermink, et al. 1984;

Blancou, et al. 1986; Cremer, et al. 1985; Earl, et al. 1986; Morrison, et ai 1986; Panicali, et al. 1983; Paoletti, et al. 1984; Smith, et al. 1983; Yewdell, etal. 1986). Second, vaccination with recombinant W can stimulate even an impaired immune system. For example, it has been shown that old mice, like elderly humans, are more susceptible to influenza despite immunization with killed vaccines. However, when vaccinated with a W recombinant containing the PR8 HA gene, they were protected from challenge and generated high levels of anti-HA Ab and PR8-specific cytotoxic T cells (Ben-Yehuda, et al. 1993). Third, the capacity of MVA or W to hold numerous inserts provides an opportunity for one construct 145 to protect against the entire range of SIV HA molecule diversity, as well as induce type- specific immunity using the NP gene. Fourth, if our vaccination/inhibition data holds true for other antigens, the MVA vector can be used to carry a variety of inserts, with repeat

vaccination as early as two months after the first recombinant regimen. Fifth, a number of studies indicate that killed vaccines do not provide as long lasting protective immunity

(Askonas, et al. 1982; Ben-Yehuda, et al. 1993), when compared to attenuated and recombinant live vaccines.

The characteristics of an ideal viral vaccine have been listed previously

(Paul 1998) and include: protection against disease, protection against infection, protection against multiple serotypes, long lasting immunity, safety for the host animal, little to no virus shedding, no reversion to virulence, and differentiation of vaccinated animals from those exposed to wildtype virus. In the present studies and others, the case is being made for MVA as an ideal vaccine vector. MVA recombinants are able to protect against disease and infection, and potentially against multiple serotj^es. MVA is safe for mammalian hosts, does not result in shedding nor reversion to virulence, and vaccinates can be distinguished from exposed animals. What is not yet determined is the duration of immunity provided.

What remains for near future testing, then, is determination of how long immune

response to MVA recombinants endures. Studies should also be conducted to establish optimum dose and time for vaccination, and whether maternal Ab interferes with immunization. In addition, it would be helpful to know which days postchallenge, of days 1 through 5, SIV can be isolated from the lungs, especially of IM-inoculated vaccinates. The lack of lesions on day 5 postchallenge in this group, the low level of virus present in the 146

nares, and the absence of clinical signs suggests that the systemic IgG response is potent and keeps viral replication in check. This hypothesis should be verified by cell culture and immunohistochemistry. Lastly, the prevalence and distribution of naturally occurring orthopoxviruses, especially in small mammalian species, should be assessed, as was done in

Norway (Sandvik, et al. 1998), and the ability of various vaccine vectors to recombine with them determined. Given that the MVA strain is replication-incompetent in mammalian cell lines, there would appear to be little to no chance for recombination to occur if it is administered by IM inoculation. Likewise, the possibility of reassortment of influenza RNA segments seems remote, given that MVA is a DNA virus residing in the cytoplasm, producing positive sense mRNA for protein production. In contrast SIV is a negative-sense RNA virus residing in the nucleus, but only of epithelial cells of the respiratory tract (Easterday &

Hinshaw 1992; Enami, et al. 1991; Murphy Webster 1996; Whittaker & Helenius 1998), making reassortment of recombinant and influenza RNA very unlikely if the MVA vaccine is given EN, and impossible if the vaccine is given IM. Although it is theoretically possible for

IN-inoculated MVA-expressed HA and NP protein to be incorporated into co-infecting SIV replication, use of the IM route, more advantageous for other reasons, such as ease of use, completely avoids this potential risk.

References

Askonas BA, McMichael AJ, Webster RG. 1982. The immune response to influenza viruses and the problem of protection against infectionBeare AS, Editor. Basic and applied influenza research. Boca Raton, FL; CRC Press, p 159-88. 147

Ben-Yehuda A, Ehleiter D, Hu A, Weksler ME. 1993. Recombinant vaccinia virus expressing the PR/8 influenza hemagglutinin gene overcomes the impaired immune response and increased susceptibility of old mice to influenza infection. J Infect Dis 168:352-7.

Bennink JR, Yewdell JW, Smith GL, Moller C, Moss B. 1984. Recombinant vaccinia virus primes and stimulates influenza haemagglutinin-specific cytotoxic T cells. Nature 31l(5986):578-9.

Blancou J, Kieny M-P, Lathe R, Lecocq JP, Pastoret PP, .Soulebot JP, Desmettres P. 1986. Oral vaccination of the fox against rabies using a live recombinant vaccinia virus. Nature 332:373.

Cremer K, Mackett M, Wohlenberg C, Notkins AL, Moss B. 1985. Vaccinia virus recombinants expressing herpes simplex virus type 1 glycoprotein D prevents latent herpes in mice. Science 228:737-40.

Earl PL, Moss B, Morrison RP, Wehrly K, Nishio J, Chesebro B. 1986. T-lymphocyte priming and protection against leukemia by vaccinia-retroviruses env gene recombinant. Science 234:728-31.

Easterday BC, Hinshaw VS. 1992. Swine influenza. Leman AD, Straw BE, Mengeling WL, D'Allaire S, Taylor DJ, eds. Diseases of swine. 7th ed. Ames, lA: Iowa State University Press, p 349-57.

Enami M, Sharma G, Benham C, Palese P. 1991. An influenza virus containing nine different RNA segments. Virology 185:291-8.

Morrison LA, Lukacher AE, Braciale'VL, Fan DP, Braciale TJ. 1986. Differences in antigen presentation to MHC class I- and class Il-restricted influenza virus specific cytolytic T lymphocyte clones. J Exp Med 163:903-21.

Murphy BR, Webster RG. 1996. Orthomjocoviruses. Fields BN, Knipe DM, Howley PM and others, eds. Fields Virology. 3rd ed. Philadelphia: Lippincott-Raven Publishers, p 1397-445.

Paul PS. International Institute for Cooperation in Animal Biologies; 1998 Jun; Ames, lA.

Sandvik T, Tryland M, Hansen H, Mehl R, Moens U, Olsvik O, Traavik T. 1998. Naturally occurring orthopoxviruses: potential for recombination with vaccine vectors. J Clin Microbiol 36(9):2542-7.

Whittaker GR, Helenius A. 1998. Nuclear import and export of viruses and virus genomes. Virology 246:1-23.

Yewdell JW, Bennink JR, Mackett M, Lefrancois L, Lyles DS, Moss B. 1986. Recognition of cloned vesicular stomatitis virus internal and external gene products by cytotoxic T lymphocytes. J Exp Med 163:1529-38.