Synthetic recombinant vaccine expressing influenza haemagglutinin epitope in Salmonella flagellin leads to partial protection in mice J u a n M c E w e n , R a p h a e l Levi, R o b e r t J. H o r w i t z a n d R u t h A r n o n *
The influenza virus haemagglutinin epitope 91-108, which is a conserved amino acid sequence in all type A H3 strains, was expressed in Salmonella flagellin, to evaluate its potential as a vaccine. For that purpose, a synthetic oligonucleotide comprising 54 bases coding for the corresponding sequence was inserted into the plasmid pLS408 and transformed into Escherichia coli JMIO1. Colonies containing the recombinant plasmid were used to transform Salmonella typhimurium LB5000 and were then transduced to a flagellin negative 'live vaccine' aroA mutant o f Salmonella dublin. Rabbits immunized either with the live recombinant S. dublin or with the flagellin isolated from it, showed significant levels o f IgG response against the synthetic peptide 91-108 as well as against the intact A / T e x a s / 7 7 influenza virus. Mice immunized with the same preparations developed influenza-specific IgG antibodies in the blood and secreted IgA antibodies in their lungs. Furthermore, these mice showed about 50% protection against challenge infection with the virus. The most successful results were achieved by intranasal immunization with the isolated recombinant flagellin, when employed without the aid o f adjuvant. Keywnrds: Influenza; recombinant; haemagglutinin; Salmonella; synthetic; epitope
INTRODUCTION Influenza virus comprises two surface antigens, the haemagglutinin (HA) and the neuraminidase (NA). The haemagglutinin is the most significant antigen in defining the serological specificity. Antibodies to the HA neutralize virus infectivity, and resistance to influenza infection was shown to correlate with serum anti-HA antibodies levels L2. Furthermore, passive transfer of immune serum proved to be protective against further infection 3. Hence, HA has been employed as an important component in the presently existing influenza vaccines, which are administered parenterally. However, various studies have suggested that resistance to respiratory viral infections is mainly mediated by antiviral IgA antibodies which are generated in respiratory mucosa, rather than in the serum 4-6. Indeed, both oral and intranasal administration of influenza antigens were shown to provoke in mice humoral IgA response in the lungs, which protected them from viral challenge 7,s. Previous studies in this laboratory, focusing on the
Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot 76 100, Israel. *To whom correspondence should be addressed. (Received 12 August 1991; revised 26 November 1991; accepted 28 November 1991) 0264-410X/92/060405-07 © 1992 Butterworth-HeinemannLtd
HA epitope 91-108 of H3 Influenza subtype, showed it to be immunogenic when coupled to tetanus toxoid 9. When this antigen was administered to mice parenterally, either in complete Freund's adjuvant (CFA) or as a conjugate with the synthetic adjuvant muramyl dipeptide (MDP), it led to partial protection against challenge infection 1°. It was therefore of interest to investigate the immune response to this epitope delivered via oral or intranasal administration. Attenuated Salmonella, which induce a broad immune response after oral administration it, offer a convenient way of presenting heterologous antigens to the immune system. In recent studies, using Salmonella as a carrier, it was possible to induce an immune response directed against several foreign antigens, including cholera toxin epitope 12, malaria antigens ta, hepatitis B surface antigen14, tetanus toxin~ ~ and streptococcal M protein~ 6. The expression of the HA epitope 91-108 in Salmonella flagellin is described here. We demonstrate a humoral immune response following oral immunization with the recombinant bacteria, or intranasal immunization with the purified chimeric flagellin. Intranasal vaccination with the hybrid flagellin also conferred partial protection against viral challenge. This is the first report, to our knowledge, on the use of a synthetic recombinant vaccine that leads to a biologically active anti-viral effect in an animal model.
Vaccine, Vol. 10, Issue 6, 1992
405
Synthetic recombinant influenza vaccine: J. McEwen et al.
MATERIALS AND METHODS
Animals Rabbits, New Zealand strain, 3 4 months old, were obtained from the Animal Breeding Center of the Weizmann Institute. Mice inbred strains Balb/c and C57BL/6 were from Olac, Blackthorn, Bicester, UK. Outbred mice strain CD1 were from the Animal Breeding Center of the Weizmann Institute of Science. All mice were used at the age of 8-12 weeks. Embryonated hen eggs, 9-11 days, were obtained from the Kfar Bilu Hatchery. Virus Influenza strain A/Texas/77 (H3N2) was grown in the allantoic cavity of 9-11-day-old embryonated eggs and used as infectious allantoic fluid. Virus was grown and purified according to standard methods 17. Titration of virus in the allantoic fluid was performed by the haemagglutinin assay using 50 #1 allantoic fluid serial dilutions and 50 #1 of 0.5% chicken red blood cells in saline. The titres were expressed as haemagglutination units (HAU). Immunization and infection procedures Rabbits were injected either with intact Salmonella intramuscularly (i.m.), or with the purified flagellin subcutaneously (s.c.). The flagellin was injected in CFA, 1 mg/animal, in 1 ml emulsion. Three booster injections were given at 3 - 4 week intervals, with half the amounts of the antigen. In the case of the flagellin, incomplete Freund's adjuvant (IFA) was used. Blood was collected 4 - 7 days after boosts. Mice were immunized either with Salmonella orally, or with the purified flagellin intranasally (i.n.). Live bacteria from an overnight culture were washed in PBS and diluted to absorbance at 600 nm (A600) of 2.5/ml. Each animal received 200 #1 orally, using animal feeding needles (Popper & Sons, NY, USA). Flagellin, 50 #g per animal in 25-50 pl PBS was administered i.n. to mice anaesthetized with ether. In immunogenicity experiments, mice immunized orally with whole bacteria received five boosts, 7 days apart. Those immunized i.n. with purified flagellin were boosted three times, 14 days apart. Boosts were given with the same amounts of antigen used for the initial immunization. The animals were killed 7 days after the last boost and antibody titres were evaluated. Lung washes and blood collection In order to assay the production of IgA (in lungs) or IgG (in serum) in the immunized mice, a procedure essentially as described by Liew et al. 4 was employed. Briefly, a cannula was inserted into the trachea, and the lungs were washed with 5 ml PBS per animal, using a three-way stopcock. The fluid obtained was centrifuged in order to remove cells and debris, concentrated and assayed for IgA. Blood accumulated in the thoracic cavity was assayed for IgG. Viral challenge and virus titration In challenge studies to assess the anti-viral activity elicited by the immunization, 7 days after the last boost mice were inoculated i.n. with 6 HAU A / T e x a s / l / 7 7 , and 3 days later the animals were killed, their lungs removed, and blood collected. All samples were stored
406
Vaccine, Vol. 10, Issue 6, 1992
at - 7 0 ° C . Immediately prior to the assay, lungs were thawed, homogenized in PBS ( 10% w/v) and centrifuged to remove debris. Virus titres were assayed by one of two methods. The first method is the allantoic-on-shell method, described by Fazekas De St Groth and White ~8. The assay was performed in U-bottomed microplates, each well containing 150 #1 standard medium, a 4 x 4 mm piece of 11 day embryonated egg shell with the chorioallantoic membrane attached, and 15/~1 of sample. The plates were incubated at 37°C with agitation for 48 h. The egg shells were then removed and 80/~1 1% RBC solution in saline was added to assay for virus presence by haemagglutination, as indicated above. The second method was whole egg titration, by injection of 100/A of the lung homogenate dilutions into the allantoic cavity of 9-11-day-old embryonated eggs. Following incubation for 48 h at 37°C and overnight at 4°C, allantoic fluid was removed. Virus presence was assayed by haemagglutination in microtitre plates containing 50/~1 allantoic fluid and 50/~1 0.5% RBC in saline.
Synthetic peptides and oligonucleotides Peptide 91-108 was synthesized by the solid phase technique in a 430A peptide synthesizer (Applied Biosystems), as described previously 1°. After cleavage from the resin, the peptide was purified on a Sephadex G-25 column. Synthetic oligonucleotides were prepared in a 380B Applied Biosystems DNA synthesizer. Preparation of recombinant bacteria The construction of the expression vector, which was generously supplied by Professor B. Stocker, was described elsewhere 12. The synthesized oligonucleotides were inserted at the E c o R V site of the plasmid pLS408, and transformed into E. coli JM101 competent cells. Colonies containing the recombinant plasmid were selected by probing them with one of the oligonucleotides labelled with 32p-ATP. Plasmids from positive colonies were purified and the insert orientation was determined using restriction analysis. The desired plasmids were used to transform Salmonella typhimurium LB5000 (a restriction negative, modification proficient non-flagellated) competent cells 19 and were then transferred to a flagellin-negative live vaccine strain (an aroA mutant) of S. dublin SL5928 by transduction using the phage P22HT105/1 int 2°'2x. The transformed S. dublin were selected for ampicillin resistance, motility under the light microscope and growth in semi-solid LB agar plates, supplemented with Oxoid nutrient broth no. 2. Selected clones were grown overnight in 21 Luzia-Bertani medium supplemented with ampicillin (50 100 #g #1-1) and the flagellin was purified by acidic cleavage, according to the technique described by Ibrahim et al. 22. ELISA A slight modification of the method described by Engvall and Perlmann 23 was used, by employing ELISA immunoplates (Nunc) instead of tubes. ELISA was used for determining the presence of the inserted peptide in the chimeric flagellin, using rabbit antibodies against the synthetic peptide 91-108, prepared previously in this laboratory 1°. The plates were pretreated with 0.2% glutaraldehyde to allow better
Synthetic recombinant influenza vaccine: J. McEwen et al.
adsorption of the peptide 24. ELISA was also used in the assay for specific lgA or IgG antibodies, respectively, in lung washes and sera. In this case the antigens adsorbed to the plate were the peptide 91-108 (1/~g/well) or intact purified A/Texas/I/77 virus (20 HAU/weI1), in 100/A PBS. Goat anti-mouse IgA antibodies conjugated to alkaline phosphatase (Sigma) and sheep antimouse IgG antibodies conjugated to fl-galactosidase (Amersham) were used as second antibodies, p-Nitrophenoyl phosphate (PNP) or o-nitrophenyl-fl-D-galactopyranoside (ONPG), 1 mg ml- 1 solutions ( 100/A/well), were used as substrates for the alkaline phosphatase and fl-galactosidase, respectively, and the plates were read at 405 nm. RESULTS
94
67
43
Expression of influenza epito~ in Salmonella flagellin A synthetic oligonucleotide comprising 54 bases, coding for amino acids 91-108 of the H3 influenza subtype, was prepared, with the codon usage corresponding to that in the sequence of Salmonella flagellin gene 25, as illustrated in Figure 1. As shown, a MluI restriction site was created that, together with the HindllI and AluI sites already present in the sequence, facilitates analysis of the product. An inverted stop codon was also chosen to prevent expression in case of insertion of the synthetic oligonucleotide in the wrong orientation (Figure 1). The synthetic oligonucleotide was inserted into the plasmid pLS408 and transformed into E. coli JM101. Colonies containing the recombinant plasmid were used to transform Salmonella typhimurium LB5000 and were then
~
I O I
Figure 2
Shift in molecular weight of the recombinant flagellin (Fla-91) when compared with the native flagellin (Fla-control), as demonstrated in SDS-PAGE (7.5%) of the purified flagellins
EcoRI
pLS408-i (6.4Kb) ORI
pl
Antigenic epitope
transduced to a flagellin negative 'live vaccine' aroA mutant of S. dublin. The recombinant bacteria carrying the 91-108 epitope were designated Salmo-91, and the flagellin purified from them was denoted Fla-91. Salmo-control and Fla-control are the terms given to the bacteria expressing flagellin coded by plasmid pLS408 and the flagellin purified from them, respectively. The presence of the influenza sequence in the flagellin protein was demonstrated by the following criteria: a shift in molecular weight as observed in SDS-PAGE (Figure 2) (the recombinant flagellin has a slightly higher molecular weight than the native one) and ELISA, using anti HA 91-108 antibodies, that demonstrated the recognition of the recombinant flagellin by these antibodies (Figure 3). Humora] immune response
Hind ~ ~ Al]U ~ III Mlu I ~istop TCT AAA GCT TTC TCT AAC TGT TAT CCT TAT GAT GTT CCG GAT TAC GCG TCT TTA Set Lys Aia Phe Set Ash Cys T y r Pro T y r Asp Val Pro Asp T y r AIa Set Leu
Fl~lure 1 Structure of pLS408 vector and the base sequence of the insert. The construction of this plasmid was described in detail elsewhere ~2. Hl-d is the flagellin gene from Salmonella munchen. The oligonucleotides sequence shown at the bottom was designed to code for epitope 91-108 of the influenza A H3 subtype HA. Codon usage was according to the sequence of the flagellin gene, with minor modifications in order to create the Mlul restriction site, and a stop codon in the inserts which are in the wrong orientation
The immunogenicity of the recombinant bacteria was first tested in rabbits, which were immunized either i.m. with the live recombinant bacteria, or s.c. with the isolated Fla-91 in the presence of CFA. Both immunizations resulted in a specific IgG response against the synthetic peptide, as well as against the intact A/Texas/1/77 virus (Figure 4a and b). In the latter case, the immunization with whole bacteria elicited higher antibody levels. Based on the results obtained in rabbits, which showed
Vaccine, Vol. 10, I ssue 6, 1992
407
Synthetic recombinant influenza vaccine: J. McEwen et al. 3
a
b
2.5
,~ v =o
•
~
¢
~.
--
~
2
c
1.5
.~
The influenza virus haemagglutinin epitope 91-108, which is a conserved amino acid sequence in all type A H3 strains, was expressed in Salmonella flagellin, to evaluate its potential as a vaccine. For that purpose, a synthetic oligonucleotide comprising 54 bases coding for the corresponding sequence was inserted into the plasmid pLS408 and transformed into Escherichia coli JMIO1. Colonies containing the recombinant plasmid were used to transform Salmonella typhimurium LB5000 and were then transduced to a flagellin negative 'live vaccine' aroA mutant o f Salmonella dublin. Rabbits immunized either with the live recombinant S. dublin or with the flagellin isolated from it, showed significant levels o f IgG response against the synthetic peptide 91-108 as well as against the intact A / T e x a s / 7 7 influenza virus. Mice immunized with the same preparations developed influenza-specific IgG antibodies in the blood and secreted IgA antibodies in their lungs. Furthermore, these mice showed about 50% protection against challenge infection with the virus. The most successful results were achieved by intranasal immunization with the isolated recombinant flagellin, when employed without the aid o f adjuvant. Keywnrds: Influenza; recombinant; haemagglutinin; Salmonella; synthetic; epitope
INTRODUCTION Influenza virus comprises two surface antigens, the haemagglutinin (HA) and the neuraminidase (NA). The haemagglutinin is the most significant antigen in defining the serological specificity. Antibodies to the HA neutralize virus infectivity, and resistance to influenza infection was shown to correlate with serum anti-HA antibodies levels L2. Furthermore, passive transfer of immune serum proved to be protective against further infection 3. Hence, HA has been employed as an important component in the presently existing influenza vaccines, which are administered parenterally. However, various studies have suggested that resistance to respiratory viral infections is mainly mediated by antiviral IgA antibodies which are generated in respiratory mucosa, rather than in the serum 4-6. Indeed, both oral and intranasal administration of influenza antigens were shown to provoke in mice humoral IgA response in the lungs, which protected them from viral challenge 7,s. Previous studies in this laboratory, focusing on the
Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot 76 100, Israel. *To whom correspondence should be addressed. (Received 12 August 1991; revised 26 November 1991; accepted 28 November 1991) 0264-410X/92/060405-07 © 1992 Butterworth-HeinemannLtd
HA epitope 91-108 of H3 Influenza subtype, showed it to be immunogenic when coupled to tetanus toxoid 9. When this antigen was administered to mice parenterally, either in complete Freund's adjuvant (CFA) or as a conjugate with the synthetic adjuvant muramyl dipeptide (MDP), it led to partial protection against challenge infection 1°. It was therefore of interest to investigate the immune response to this epitope delivered via oral or intranasal administration. Attenuated Salmonella, which induce a broad immune response after oral administration it, offer a convenient way of presenting heterologous antigens to the immune system. In recent studies, using Salmonella as a carrier, it was possible to induce an immune response directed against several foreign antigens, including cholera toxin epitope 12, malaria antigens ta, hepatitis B surface antigen14, tetanus toxin~ ~ and streptococcal M protein~ 6. The expression of the HA epitope 91-108 in Salmonella flagellin is described here. We demonstrate a humoral immune response following oral immunization with the recombinant bacteria, or intranasal immunization with the purified chimeric flagellin. Intranasal vaccination with the hybrid flagellin also conferred partial protection against viral challenge. This is the first report, to our knowledge, on the use of a synthetic recombinant vaccine that leads to a biologically active anti-viral effect in an animal model.
Vaccine, Vol. 10, Issue 6, 1992
405
Synthetic recombinant influenza vaccine: J. McEwen et al.
MATERIALS AND METHODS
Animals Rabbits, New Zealand strain, 3 4 months old, were obtained from the Animal Breeding Center of the Weizmann Institute. Mice inbred strains Balb/c and C57BL/6 were from Olac, Blackthorn, Bicester, UK. Outbred mice strain CD1 were from the Animal Breeding Center of the Weizmann Institute of Science. All mice were used at the age of 8-12 weeks. Embryonated hen eggs, 9-11 days, were obtained from the Kfar Bilu Hatchery. Virus Influenza strain A/Texas/77 (H3N2) was grown in the allantoic cavity of 9-11-day-old embryonated eggs and used as infectious allantoic fluid. Virus was grown and purified according to standard methods 17. Titration of virus in the allantoic fluid was performed by the haemagglutinin assay using 50 #1 allantoic fluid serial dilutions and 50 #1 of 0.5% chicken red blood cells in saline. The titres were expressed as haemagglutination units (HAU). Immunization and infection procedures Rabbits were injected either with intact Salmonella intramuscularly (i.m.), or with the purified flagellin subcutaneously (s.c.). The flagellin was injected in CFA, 1 mg/animal, in 1 ml emulsion. Three booster injections were given at 3 - 4 week intervals, with half the amounts of the antigen. In the case of the flagellin, incomplete Freund's adjuvant (IFA) was used. Blood was collected 4 - 7 days after boosts. Mice were immunized either with Salmonella orally, or with the purified flagellin intranasally (i.n.). Live bacteria from an overnight culture were washed in PBS and diluted to absorbance at 600 nm (A600) of 2.5/ml. Each animal received 200 #1 orally, using animal feeding needles (Popper & Sons, NY, USA). Flagellin, 50 #g per animal in 25-50 pl PBS was administered i.n. to mice anaesthetized with ether. In immunogenicity experiments, mice immunized orally with whole bacteria received five boosts, 7 days apart. Those immunized i.n. with purified flagellin were boosted three times, 14 days apart. Boosts were given with the same amounts of antigen used for the initial immunization. The animals were killed 7 days after the last boost and antibody titres were evaluated. Lung washes and blood collection In order to assay the production of IgA (in lungs) or IgG (in serum) in the immunized mice, a procedure essentially as described by Liew et al. 4 was employed. Briefly, a cannula was inserted into the trachea, and the lungs were washed with 5 ml PBS per animal, using a three-way stopcock. The fluid obtained was centrifuged in order to remove cells and debris, concentrated and assayed for IgA. Blood accumulated in the thoracic cavity was assayed for IgG. Viral challenge and virus titration In challenge studies to assess the anti-viral activity elicited by the immunization, 7 days after the last boost mice were inoculated i.n. with 6 HAU A / T e x a s / l / 7 7 , and 3 days later the animals were killed, their lungs removed, and blood collected. All samples were stored
406
Vaccine, Vol. 10, Issue 6, 1992
at - 7 0 ° C . Immediately prior to the assay, lungs were thawed, homogenized in PBS ( 10% w/v) and centrifuged to remove debris. Virus titres were assayed by one of two methods. The first method is the allantoic-on-shell method, described by Fazekas De St Groth and White ~8. The assay was performed in U-bottomed microplates, each well containing 150 #1 standard medium, a 4 x 4 mm piece of 11 day embryonated egg shell with the chorioallantoic membrane attached, and 15/~1 of sample. The plates were incubated at 37°C with agitation for 48 h. The egg shells were then removed and 80/~1 1% RBC solution in saline was added to assay for virus presence by haemagglutination, as indicated above. The second method was whole egg titration, by injection of 100/A of the lung homogenate dilutions into the allantoic cavity of 9-11-day-old embryonated eggs. Following incubation for 48 h at 37°C and overnight at 4°C, allantoic fluid was removed. Virus presence was assayed by haemagglutination in microtitre plates containing 50/~1 allantoic fluid and 50/~1 0.5% RBC in saline.
Synthetic peptides and oligonucleotides Peptide 91-108 was synthesized by the solid phase technique in a 430A peptide synthesizer (Applied Biosystems), as described previously 1°. After cleavage from the resin, the peptide was purified on a Sephadex G-25 column. Synthetic oligonucleotides were prepared in a 380B Applied Biosystems DNA synthesizer. Preparation of recombinant bacteria The construction of the expression vector, which was generously supplied by Professor B. Stocker, was described elsewhere 12. The synthesized oligonucleotides were inserted at the E c o R V site of the plasmid pLS408, and transformed into E. coli JM101 competent cells. Colonies containing the recombinant plasmid were selected by probing them with one of the oligonucleotides labelled with 32p-ATP. Plasmids from positive colonies were purified and the insert orientation was determined using restriction analysis. The desired plasmids were used to transform Salmonella typhimurium LB5000 (a restriction negative, modification proficient non-flagellated) competent cells 19 and were then transferred to a flagellin-negative live vaccine strain (an aroA mutant) of S. dublin SL5928 by transduction using the phage P22HT105/1 int 2°'2x. The transformed S. dublin were selected for ampicillin resistance, motility under the light microscope and growth in semi-solid LB agar plates, supplemented with Oxoid nutrient broth no. 2. Selected clones were grown overnight in 21 Luzia-Bertani medium supplemented with ampicillin (50 100 #g #1-1) and the flagellin was purified by acidic cleavage, according to the technique described by Ibrahim et al. 22. ELISA A slight modification of the method described by Engvall and Perlmann 23 was used, by employing ELISA immunoplates (Nunc) instead of tubes. ELISA was used for determining the presence of the inserted peptide in the chimeric flagellin, using rabbit antibodies against the synthetic peptide 91-108, prepared previously in this laboratory 1°. The plates were pretreated with 0.2% glutaraldehyde to allow better
Synthetic recombinant influenza vaccine: J. McEwen et al.
adsorption of the peptide 24. ELISA was also used in the assay for specific lgA or IgG antibodies, respectively, in lung washes and sera. In this case the antigens adsorbed to the plate were the peptide 91-108 (1/~g/well) or intact purified A/Texas/I/77 virus (20 HAU/weI1), in 100/A PBS. Goat anti-mouse IgA antibodies conjugated to alkaline phosphatase (Sigma) and sheep antimouse IgG antibodies conjugated to fl-galactosidase (Amersham) were used as second antibodies, p-Nitrophenoyl phosphate (PNP) or o-nitrophenyl-fl-D-galactopyranoside (ONPG), 1 mg ml- 1 solutions ( 100/A/well), were used as substrates for the alkaline phosphatase and fl-galactosidase, respectively, and the plates were read at 405 nm. RESULTS
94
67
43
Expression of influenza epito~ in Salmonella flagellin A synthetic oligonucleotide comprising 54 bases, coding for amino acids 91-108 of the H3 influenza subtype, was prepared, with the codon usage corresponding to that in the sequence of Salmonella flagellin gene 25, as illustrated in Figure 1. As shown, a MluI restriction site was created that, together with the HindllI and AluI sites already present in the sequence, facilitates analysis of the product. An inverted stop codon was also chosen to prevent expression in case of insertion of the synthetic oligonucleotide in the wrong orientation (Figure 1). The synthetic oligonucleotide was inserted into the plasmid pLS408 and transformed into E. coli JM101. Colonies containing the recombinant plasmid were used to transform Salmonella typhimurium LB5000 and were then
~
I O I
Figure 2
Shift in molecular weight of the recombinant flagellin (Fla-91) when compared with the native flagellin (Fla-control), as demonstrated in SDS-PAGE (7.5%) of the purified flagellins
EcoRI
pLS408-i (6.4Kb) ORI
pl
Antigenic epitope
transduced to a flagellin negative 'live vaccine' aroA mutant of S. dublin. The recombinant bacteria carrying the 91-108 epitope were designated Salmo-91, and the flagellin purified from them was denoted Fla-91. Salmo-control and Fla-control are the terms given to the bacteria expressing flagellin coded by plasmid pLS408 and the flagellin purified from them, respectively. The presence of the influenza sequence in the flagellin protein was demonstrated by the following criteria: a shift in molecular weight as observed in SDS-PAGE (Figure 2) (the recombinant flagellin has a slightly higher molecular weight than the native one) and ELISA, using anti HA 91-108 antibodies, that demonstrated the recognition of the recombinant flagellin by these antibodies (Figure 3). Humora] immune response
Hind ~ ~ Al]U ~ III Mlu I ~istop TCT AAA GCT TTC TCT AAC TGT TAT CCT TAT GAT GTT CCG GAT TAC GCG TCT TTA Set Lys Aia Phe Set Ash Cys T y r Pro T y r Asp Val Pro Asp T y r AIa Set Leu
Fl~lure 1 Structure of pLS408 vector and the base sequence of the insert. The construction of this plasmid was described in detail elsewhere ~2. Hl-d is the flagellin gene from Salmonella munchen. The oligonucleotides sequence shown at the bottom was designed to code for epitope 91-108 of the influenza A H3 subtype HA. Codon usage was according to the sequence of the flagellin gene, with minor modifications in order to create the Mlul restriction site, and a stop codon in the inserts which are in the wrong orientation
The immunogenicity of the recombinant bacteria was first tested in rabbits, which were immunized either i.m. with the live recombinant bacteria, or s.c. with the isolated Fla-91 in the presence of CFA. Both immunizations resulted in a specific IgG response against the synthetic peptide, as well as against the intact A/Texas/1/77 virus (Figure 4a and b). In the latter case, the immunization with whole bacteria elicited higher antibody levels. Based on the results obtained in rabbits, which showed
Vaccine, Vol. 10, I ssue 6, 1992
407
Synthetic recombinant influenza vaccine: J. McEwen et al. 3
a
b
2.5
,~ v =o
•
~
¢
~.
--
~
2
c
1.5
.~
