- Email: [email protected]
89 (H3N8) — A new equine influenza virus
Pc pers Efficacy of equine influenza vaccines for protection against A/Equine/Jilin/89 ( H 3 N 8 ) - a new equine influenza virus Robert G. Webster* and Timothy L. Thomas .4 new H3N8 equine influenza virus [A/Equine/Jilin/1/89 (Eq/Jilin)] appeared in Northeastern China in 1989 and caused high mortality in horses; the available evidence indicates that it has not yet spread outside this region of the world. Serological analysis with postinfection ferret sera in haemagglutination inhibition (HI) tests confirmed that Eq/Jilin is antigenically distinct from H3N8 equine influenza viruses isolated between 1963 and 1991 and also showed that a current equine influenza virus [A/Equine/Alaska/ 1/91 ( H3N8 ) ] had undergone antigenic drift. In the present study we determine if vaccine against a recent H3N8 influenza virus [A/Equine/Kentucky/1277/90 ( Eq/ Kentucky) ] that was standardized for haemagglutinin content will protect mice against lethal challenge with the new H3N8 influenza virusfrom China. Complete protection is defined as prevention of virus replication in the lungs of mice 3 days after challenge. High doses of Eq/Kentucky vaccine in aqueous suspension (0.5-5.0 #g HA per dose) provided minimal protection against Eq/Jilin challenge as judged by virus titres in the lungs of vaccinated animals. Eq/Kentucky vaccine in adjuvant (1.0-5.0#9 HA per dose) did provide complete protection against challenge with Eq/ Jilin in mice. Eq/ Jilin vaccine in aqueous suspension induced complete protection of mice against challenge with Eq/Kentucky at doses from 0.5 to 5 #9 HA and in adjuvant doses of Eq/Jilin from 0.1-5.0 I~g HA were efficacious. Homologous protection against Eq/ Jilin or Eq/ Kentucky was induced by doses of vaccine from 0.5-5.0 i~9 HA per dose in aqueous suspension and from 0.01-5.0 #9 HA per dose in adjuvant. Lower doses of vaccine (0.01-0.1 #9 HA)failed to induce detectable HI antibodies yet the mice were protected," ELISA and neutralizing antibodies correlated with protection. Studies showed that commercially available equine influenza vaccines provided minimal cross-protection with Eq / Jilin virus but did prevent mortality. The present studies indicate that cross-protection can be induced against the new H3N8 equine influenza virus from China with vaccine to the currently circulating equine H3N8 viruses but high doses of vaccine are needed," commercially available vaccines do not contain sufficient antigen to induce complete protection in mice. Keywords:Equine influenza virus; cross-protection; adjuvant
In 1989 an H3N8 influenza virus caused a severe outbreak of respiratory disease in horses in Jilin and Heilongjiang provinces of Northeastern China 1. Morbidity was 81% and mortality was as high as 20% in some herds. The causative agent has been characterized as a new H3N8 influenza virus that derived its entire genome from an influenza virus in an avian species and the prototype strain is A/Equine/Jilin/1/89 (H3N8) ( Eq/Jilin)X. Since 1989, influenza viruses closely related to the Eq/Jilin strain caused a second outbreak of disease in horses in
Department of Virology and Molecular Biology, St Jude Children's Research Hospital, 332 North Lauderdale, PO Box 318, Memphis, TN 38101, USA. *To whom correspondence should be addressed. (Received 30 June 1992; revised 28 August 1992; accepted 28 August 1992) 0264-410X/93/10/0987--07 © 1993 Butterworth-HeinemannLtd
April 1990 in Heilongjiang province with 48% morbidity and no mortality. No influenza was detected in horses in Northern China or adjacent areas in 1991. The question is whether this virus will appear in horses elsewhere in the world and spread or whether it will behave like the seal influenza virus that caused high mortality for 1 year in seals and then disappeared 2. Since the density of horses is relatively high in adjacent areas of Mongolia, the possibility still exists for the new equine strain to spread. In the present study we determine the efficacy of vaccine to an equine H3N8 strain currently circulating in USA to protect against challenge with the new H3N8 isolate from China. We used a mouse model system for these studies and compared the efficacy of vaccine administered in aqueous suspension and in adjuvant. The studies show that vaccines to A / E q u i n e / K e n t u c k y / 1 2 7 7 /
Vaccine, Vol. 11, Issue 10, 1993
987
Equine influenza vaccines: R.G. Webster and T.L. Thomas
90 (H3N8) (Eq/Kentucky) that are highly efficacious at inducing homologous protection are less efficacious against challenge with Eq/Jilin influenza virus. In contrast, vaccines to Eq/Jilin were efficacious against both Eq/Jilin and Eq/Kentucky. MATERIALS AND METHODS
Viruses A/Equine/Jilin/1/89 (H3N8) (Eq/Jilin) and A/Equine/ Kentucky/1277/90 (H3N8) (Eq/Kentucky) were used in these studies. The viruses were grown in 11-day-old embryonated chicken eggs and were purified by velocity sedimentation through 25-70% sucrose gradients in a Beckman SW28 rotor 3. All studies with Eq/Jilin were done in a biosafety level 3 containment laboratory that was approved for such use by the United States Department of Agriculture.
Virus proteins on polyacrylamide gels were stained with Coomassie brilliant blue and quantified using a Bio Image Visage 110 image analyser (Millipore Corp., Bradford, MA). After standardization by estimation of the HA content the influenza virus was inactivated with 0.025% formalin at 4°C for 3 days. The vaccine was tested for residual infectivity by two blind passages in embryonated chicken eggs and the inactivated vaccine was diluted to give 5.0, 1.0, 0.5 and 0.01 #g HA ml- 1 in phosphate-buffered saline.
Mouse adaptation of equine viruses The Eq/Jilin strain rapidly adapted to mice and was 100% lethal by the third mouse-to-mouse passage 1. The Eq/Kentucky influenza virus was passaged in mice multiple times but the virus caused no mortality. Viruses from the fourth and fifth passages were used in the experiments.
Antisera and specific antibodies
Vaccination and challenge of mice
Antisera were prepared in ferrets to the equine strains as recommended4.
Groups of 12-week-old female Balb/c mice were inoculated intraperitoneally with 1.0 ml of the standardized vaccines. The vaccine was suspended either in phosphate-buffered saline or in 10% Havlogen C in phosphate-buffered saline (an adjuvant used in a commercial equine influenza vaccine and kindly provided by Dr K.K. Brown, Miles Inc.). Four weeks after vaccination, mice were anaesthetized with metophane and inoculated intranasally with 10'~ mouse infectious doses (MID)5o/50 #1 of either mouseadapted Eq/Jilin or Eq/Kentucky. Three days after challenge, four mice per group were killed, and their lungs homogenized and titrated in embryonated chicken eggs for virus infectivity. The remaining mice in each group were observed for 10 days for mortality and were bled for serological studies.
Serological tests HA titrations and haemagglutination-inhibition (HI) tests were performed in microtitre plates with receptordestroying enzyme-treated sera 4. ELISA tests were done as described 5. Pre-immune, 3-week postvaccination, and 10-day postchallenge mouse serum samples were assayed for antibody by each procedure. Neutralization tests were done in embryonated chicken eggs using 100 egg infectious doses (EIDso) of virus.
Preparation of non-standardized inactivated vaccines Purified influenza virus at a concentration of 40000 HA ml-~ was treated with 0.025% formalin at 4°C for 3 days. Mice immunized with high doses of Eq/Jilin and Eq/Kentucky (10000 HA in Freund's adjuvant) were bled and challenged with the respective viruses.
Preparation of standardized inactivated vaccines The protein content of purified influenza virus was estimated using a modification of the Lowry method 6. The haemagglutinin content of the purified virus was determined before formalin treatment using the SDSPAGE system of Laemmli7, modified by Oxford et al. s.
Table 1
RESULTS
Serological relationships among equine influenza viruses Previous studies with chicken antisera established that Eq/Jilin influenza virus is more closely related to avian strains than to human or equine H3 influenza viruses 1. To determine if Eq/Jilin is related to any equine H3 viruses, we tested a wider range of equine strains using ferret antisera in haemagglutination inhibition (HI) assays (Table 1 ). Eq/Jilin was antigenically distinct from
Serological relationships among equine influenza viruses HI antibody titres with ferret antiserum to:
Virus
EQ/Miami/63 F100
EQ/Urg/63 F142
EQ/KY/76 F145
EQ/Font/79 F138
EQ/Alas/91 F101
EQ/Jil/89 F200
EQ/Miami/1/63 EQ/Uruguay/1/63 EQ/KY/1/76 EQ/Fontainebleau/1/79 EQ/KY/1/81 EQ/TN / 5/86 EQ/KY/1277/90 EQ/Alaska/1/91 EQ/Jilin/1/89
160" 40 160 80 160 40 160 < 10 < 10
< 10 40 80 40 20 40 80 10 < 10
< 10 10 320 80 160 160 320 20 40
< 10 < 10 80 40 10 40 80 < 10 < 10
< 10 10 10 20 20 20 10 80 10
< 10-10 10 20 < 10-10 < 10-10 10 10 10 320
aHaemagglutination inhibition titres are expressed as the reciprocal of the highest antiserum dilution inhibiting 4 haemagglutinating units of virus
988
Vaccine, Vol. 11, Issue 10, 1993
Equine influenza vaccines: R.G. Webster and T.L. Thomas
the tested equine influenza viruses isolated between 1963 and 1991. The only detectable cross-reaction was with Eq/Kentucky/76 and this was approximately 100-fold different (320 versus 20; and 320 versus 40). A point of interest is that the Equine/Alaska/1/91 (H3N8) virus is antigenically distinguishable using ferret antiserum from equine influenza viruses isolated from the USA between 1976 and 1990, suggesting that antigenic drift is occurring in equine influenza viruses. These studies with a wider range of equine influenza viruses and ferret antisera confirm the earlier finding 1, obtained with chicken antiserum, that Eq/Jilin is antigenically different from influenza viruses that have circulated in horses since 1963. Will high doses of equine influenza vaccines provide cross-protection against the new Eq/Jilin strain? A crucial question concerning the Eq/Jilin influenza virus is whether vaccines to the H3N8 influenza virus currently circulating in horses in the world will provide protection against the Eq/Jilin strain. Ideally this should be studied in horses but quarantine problems preclude this possibility. Since the Eq/Jilin strain is lethal in mice it provides a satisfactory mouse model system. Eq/Kentucky was adapted to mice and served as a representative of strains currently circulating in the USA. The Eq/Jilin strain grew to high titres in the lungs and killed all the non-vaccinated mice in 10 days; the Eq/Kentucky strain replicated in the mouse lungs but caused no mortality (Table 2). Each of the vaccines induced high levels of HI antibody before challenge but the antibody reacted only with the vaccinating virus; there was no serological cross-reaction between Eq/Kentucky and Eq/Jilin. The vaccines provided complete protection against the homologous strain - no virus was detected in the lungs of the challenged mice and no mice showed signs of disease. In cross-protection studies, neither vaccine provided complete protection but reduced the level of virus replication by three orders of magnitude and both vaccines prevented mortality after Eq/Jilin challenge. HI studies on postchallenge sera showed no detectable cross-reactivity between the Kentucky and Jilin strains. These studies indicate that, despite the absence of detectable serological cross-reactivity in HI tests between the Eq/Jilin and Eq/Kentucky strains, high doses of
vaccine will provide partial cross-protection between these H3N8 viruses. Dose response of equine influenza vaccines administered to mice in aqueous suspension The above studies provide no information on the dose of vaccine required to provide protection against homologous or heterologous challenge, and the vaccines were administered in Freund's complete adjuvant which is unacceptable in horses. Vaccines standardized in terms of haemagglutinin content were administered in aqueous suspension to mice to establish the dose response required to induce protective immunity. Complete protection was taken as the level of vaccine that prevented virus replication in the lungs of mice 3 days after challenge (Tables 3 and 4). Doses of ~>0.1/~g of Eq/Kentucky induced homologous protection; in contrast, Eq/Jilin required doses >~0.5#g for homologous protection (Tables 3 and 4 ). The lowest dose of Eq/Kentucky tested (0.01 #g HA) induced partial homologous protection by reducing the level of virus in the lungs of mice. Cross-protection differed with the vaccine strain; Eq/Kentucky vaccine provided only limited protection against challenge with Eq/Jilin even at the highest doses tested (5/~g HA) (Table 3 ). Although the Eq! Kentucky vaccine did not prevent replication of Eq/Jilin in the lungs at doses from 0.5-5.0 #g HA, it did protect against the lethality of Eq/Jilin (results not shown). On the other hand, Eq/Jilin vaccine at 0.5 #g HA gave complete protection against Eq/Kentucky challenge (Table 4). Serological responses were measured by HI, ELISA and by neutralization of infectivity. The HI assay detected antibody at high doses of antigen (0.5-5.0/~g) but not at low doses of antigen (0,1 #g) (Table 3) and did not correlate with protection. In contrast, ELISA and neutralization assays correlated with protection. Eq/Kentucky vaccine failed to induce cross-reactive antibody whereas high doses of Eq/Jilin vaccines did induce cross-reactive antibodies that were detectable in ELISA and neutralization assays. These studies indicate that equine influenza vaccines ~to the currently circulating strains (Eq/Kentucky) administered in aqueous suspension do not provide complete protection against challenge with the new Eq/Jilin strains. There was reduction in the levels of virus that replicated in the lungs of Eq/Kentucky vaccinated mice after Eq/Jilin challenge.
Table 2 Will high doses of equine influenza vaccines provide cross-protection against the new Equine/Jilin strain? Reactivity in mice
HI antibody to the following viruses:
Vaccine
Challenge virus
Sick/dead/total Day 10 PI
Virus titre (Ioglo) lung day 3 PI
Prechallenge EQ/KY/90
Prechallenge EQ/Jilin/89
Postchallenge EQ/KY/90
Postchallenge EQ/Jilin/89
None EQ/KY EQ/Jilin None EQ/KY EQ/Jilin
EO/Jilin EQ/Jilin EQ/Jilin EQ/KY EQ/KY EQ/KY
6/6/6 0/0/7 0/0/7 0/0/6 0/0/7 0/0/6
>6.5 3.6 (0-6.5) 0 5.75 (5.5-6.3) 0 1.2 (0-4.75)
< 10 328 (0-1280) < 10 < 10 222 (40-1280) <10
< 10 < 10 316 (40-1280) < 10 < 10 225 (40-1280)
Dead 1440 < 10 < 10 5260 <10
Dead < 10 240 < 10 40 160
Groups of ten Balb/c mice were immunized with 10000 HA of formalin-inactivated viruses in Freund's complete adjuvant by the intraperitoneal route. Mice were bled 3 weeks later and were challenged with 10~ mouse infectious doses 4 weeks after initial vaccination. Four mice from each group were killed 3 days postinfection (PI) and assayed for infectious virus. The mice were observed for 10 days for morbidity and mortality and bled 10 days postchallenge
Vaccine, Vol. 11, Issue 10, 1993
989
Equine influenza vaccines: R.G. Webster and T.L. Thomas Table 3
Dose response of Equine/Kentucky and Equine/Jilin vaccines in aqueous suspension in mice
Vaccine: EQ/KY/90 aqueous dose (#g HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Antibody titres to Eq/Kentucky
Challenge virus
Virus titres in mouse lungs day 3 PI
HI
EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL
0 0 0 0 1.9 2.8 4.3 5.0 5.1 7.0 6.8 6.6
8 5 7 < 10 < 10 <10 10 5 3 <10 <10 < 10
Prechallenge
Antibody titres to Eq/Jilin
Postchallenge
ELISA"
Neut.
HI
ELISA"
HI
33 10 7
400 40 <10 ~>560 125 < 10
45 55 45 40 10 <10 90 110 210 13 <10 Dead
100 100 55 33 255 155 55 Dead
<10 <10 < 10 < 10 < 10 <10 <10 <10 <10 <10 <10 < 10
< 25 7 < <
Postchallenge
Prechallenge ELISA = < < < < < < < <
Neut.
HI
< 10 < 10 <10 <10 <10 < 10
<10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10
<
Dead
Dead
ELISA"
< < <
_ _
33 25 7
-, Not tested; PI, postinfection aELISA titres x 10 -3 ( < = < 1000)
Table 4
Dose response of Equine/Kentucky and Equine/Jitin vaccines in aqueous suspension in mice
Vaccine: EQ/Jil/89 aqueous dose (/Jg HA)
Challenge virus
Virus titres in mouse lungs day 3 PI
Antibody titres to Eq/Jilin
5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY
0 0 0 3.1 6.9 6.6 0 0 0 0.8 2.5 2.8
Prechallenge HI
< < <
< <
22 15 12 10 10 10 27 22 18 2.5 10 10
Antibody titres to Eq/Kentucky
Postchaltenge
ELISA"
Neut.
HI
ELISA"
HI
10
~>560 400 < 10 400 40 < 10
20 15 20 20 45 Dead 40 25 25 < 10 5 < 10
33 25 33 Dead 18 25 7 <
<10
< < < 10 7 < <
Postchallenge
Prechallenge ELISA"
Neut.
HI
3
18 18 <10 12 12 <10
<10 <10 <10 <10 <10
<10 <10 <10 <10 < 10 <10 <10 <10 <10 <10 <10
< < < < < < <
ELISA"
10 33 33
Dead
Dead
5 <10 7.5 <10 <10 <10
55 100 55 33
-, Not tested; PI, postinfection aELISA titres x l 0 -3 ( < = <1000)
Dose response to equine influenza vaccines administered in Haviogen C adjuvant The preliminary studies (Table 2) used vaccine in Freund's complete adjuvant to induce partial crossprotection between Eq/Kentucky vaccine and Eq/Jilin challenge. Since Freund's complete adjuvant is unacceptable in horses we tested Havlogen C adjuvant, which has been used in commercial equine influenza vaccines. We determined the dose of antigen in adjuvant required to induce protective immunity and compared it with the studies on vaccine administered in aqueous suspension. Doses of Eq/Kentucky vaccine in adjuvant from 0.01 to 5/~g HA induced homologous protection (Tables 5 and 6). The adjuvant produced at least a tenfold increase in the immunogenicity of the vaccine administered in aqueous suspension. High doses of Eq/Kentucky in adjuvant ( 1.0, 5.0 #g HA ) provided cross-protection from Eq/Jilin challenge and lower doses of Eq/Jilin in adjuvant (0.1-5.0#g HA) provided cross-protection from Eq/Kentucky (Tables 5 and 6). HI antibody titres did not correlate with protection but ELISA titres did correlate both with protection and neutralizing antibody. These studies indicate that administration of equine
990
V a c c i n e , Vol. 11, I s s u e 10, 1993
influenza vaccines in adjuvant increases the immunogenicity of the vaccine both for homologous and heterologous challenge.
Will commercially available equine influenza vaccine protect mice against challenge with equine Jilin virus? Groups of Balb/c mice were immunized with commercial equine influenza virus vaccines available in the USA and were challenged with either Eq/Kentucky or Eq/Jilin (Table 7). Each vaccine provided complete protection against Eq/Kentucky and gave partial protection against Eq/Jilin. The levels of Eq/Jilin in the lungs of vaccinated mice were reduced and the animals were protected from lethal infection. No HI antibody was detected in mice before challenge but ELISA showed detectable levels of antibody. This study shows that commercially available equine influenza vaccines protect against challenge with viruses related to Eq/Kentucky but provided incomplete protection against challenge with Eq/Jilin. The absence of detectable levels of HI antibodies induced before challenge suggests that the dose of HA antigen in the vaccines is in the range 0.01-0.1/~g HA per dose.
Equine influenza vaccines: R.G. Webster and T.L. Thomas Table 5
Dose response of Equine/Kentucky and Equine/Jilin vaccines in adjuvant in mice
Vaccine: EQ/KY/90 adjuvant dose (/~g HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Antibody titres to Eq/Jilin
Antibody titres to Eq/Kentucky
Challenge virus
Virus titres in mouse lungs day 3 P!
HI
ELISA =
Neut.
HI
KY/90 KY/90 KY/90 KY/90 KY/90 KY/90 JIL/89 JIL/89 JIL/89 JIL/89 JIL/89 JIL/89
0 0 0 0 0 2.8 0 0 1.8 2.9 6.0 6.6
32 13 43 30 < 10 < 10 40 30 27 37 < 10 < 10
178 100 70 < 333 178 33 <
>~560 ~>560 >~560 >~560 -
55 120 100 60 20 < 10 240 80 130 120 60 Dead
Postchallenge
Prechallenge
ELISA"
Prechallenge HI
ELISA"
Postchallenge Neut.
HI
ELISA °
-
< 1 0
-
12
< 1 0
-
-
< 1 0
-
-
< 1 0
-
<10 <10 <10 <10
10 7
<10 <10 <10 <10
10 14
333
255 100 33
< <
-
<10
-
-
< 1 0
-
< 10 < 10 <10 < 10
10 3
778
778 100 Dead
<10 -
<10 < 10 -
< <
< <
< 1 0
<10 < 10 < 10 <10 Dead
-
78 55 < Dead
-, Not tested; PI, postinfection =ELISA titres × 10 -3 ( < = < 1000)
Table 6 Vaccine: EQ/Jil/89 adjuvant dose (/zg HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Dose response of Equine/Kentucky and Equine/Jilin vaccines in adjuvant in mice Antibody titres to EQ/Jilin
Challenge virus
Virus titres in mouse lungs day 3 PI
HI
EQ/JiL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY
0 0 0 0 0 6.6 0 0 0 0 0.8 2.8
80 100 107 60 3 < 10 88 127 107 40 4 < 10
Prechallenge
Antibody titres to EQ/Kentucky
Postchallenge
Prechallenge
ELISA °
Neut.
HI
ELISA"
HI
33 78 10
~>560 -100 >~560 ~>560 -
100 200 267 80 60 Dead 107 150 150 80 120 < 10
78 78 33 Dead 100 25 48 <
< 10 < 10 <10 < 10 < 10 < 10 <10 <10 <10 < 10 < 10 < 10
< 100 25 25 <
ELISA" 33 25 10 < 55 25 25 <
Postchallenge Neut.
HI
ELISA °
40 20 20 60 -
< 10 < 10 <10 < 10 < 10 Dead <10 125 7.5 < 10 < 10 < 10
78 33 33 Dead 255 155 178 33
-, Not tested; PI, postinfection "ELISA titres x 10 -3 ( < = < 1000)
Table 7
Do commercially available equine influenza vaccines protect mice against challenge with Equine/Jilin influenza virus?
Virus titres (Ioglo) in the lungs of mice after challenge with the following viruses: Commercial equine influenza vaccines"
Eq/KY/1277/90
A
<0.5
B C None
<0.5 NT 3.4 (3.25-3.5)
Antibody titres to Eq/Kentucky when challenged with Eq/Kentucky Prechallenge
Postchallenge
Antibody titres to Eq/Jilin when challenged with Eq/Jilin Prechallenge
Postchallenge
Eq/Jilin/1/89
HI
ELISA
HI
ELISA
HI
ELISA
HI
ELISA
4.75(4.25-5.75) 5.1 (4.75-5.5) 4.0 (3.5-4.5) 7.1 (6.75-7.25)
<10 <10 NT <10
1000 1000 NT <1000
20 10 NT <10
100000 100000 100 000 10000
<10 <10 < 10 <10
1000 1000 1000 <1000
<10 80 20 Died
100000 100000 10 000
Groups of six mice were vaccinated intraperitoneally with 0.2 ml of commercially available equine influenza vaccines; 4 weeks after vaccination the mice were challenged with approximately 10" mouse infectious doses of virus in 50 ~1 by the intranasal route. Three days after challenge the mice were killed and the virus titres in the lungs determined in embryonated chicken eggs. Serological tests were done on the remaining mice as given in Materials and methods NT, not tested "These vaccines contain equine 1 (H7N7) and equine 2 (H3N8) viruses but neither the antigen content nor the strains included are revealed
V a c c i n e , Vol. 11, I s s u e 10, 1993
991
Equine influenza vaccines: R.G. Webster and T.L. Thomas
DISCUSSION The appearance of a new equine influenza virus of the H3N8 subtype in horses i that is genetically closely related to avian H3N8 influenza viruses suggests that this particular subtype has the potential to replicate and cause serious disease problems in horses. Phylogenetic analysis of H3 haemagglutinin genes 9 indicates that the H3N8 equine-2 influenza viruses currently circulating in horses also originated from avian species. Thus we might expect that the Eq/Jilin (H3NS) will also have the potential to spread in horses. At this time we do not know if the new H3N8 Eq/Jilin virus will spread to susceptible horses elsewhere in the world. If it does, we need to know the efficacy of available equine vaccines. The appearance of influenza virus in horses in Alaska in 1991 raised the possibility that the Eq/Jilin H3N8 strain had spread; the Alaska/I/91 (H3N8) strain was associated with 100% morbidity and low mortality in some herds 1°. Serological analysis (Table 1 ) and partial sequencing of each gene established that the Alaskan isolate belongs to the equine H3N8 lineage circulating in the USA in 1989-1990. However, it is noteworthy that the Alaskan strain has shown antigenic drift, being at least fourfold different by HI from 1989-1990 equine H3N8 virus in the USA. If the new H3N8 influenza virus spreads outside Northeastern China, will vaccines currently available provide protection? This is the central question addressed in the present study using a mouse model.
Homologous protection These studies indicate that a single dose of aqueous equine vaccine to Eq/Kentucky or Eq/Jilin (0.1/~g of Eq/Kentucky and 0.5/~g of Eq/Jilin) induces complete homologous protection and that tenfold lower doses are efficacious if administered in adjuvant. Serological responses as measured by HI assays did not correlate witl~ protection but ELISA and neutralization assays showed good correlation with protection.
Cross-protection High doses of Eq/Kentucky vaccine administered in aqueous suspension gave reductions in the levels of Eq/Jilin virus in the lungs of mice (Table 3) and the mice did not die. Eq/Jilin vaccine gave significantly more cross-protection against Eq/Kentucky challenge (Table 4). Cross-protection occurred in the absence of detectable cross-reactive HI antibodies but ELISA and neutralization tests correlated with protection. Administration of high doses of either vaccine in adjuvant (1.0/~g HA) provided complete protection against heterologous challenge. No virus was detected in the lungs of infected mice and there was no mortality. Thus high doses of equine influenza virus vaccines administered in adjuvant will provide cross-protection between Eq/Kentucky and Eq/Jilin. Studies in horses 1~ showed that aqueous vaccine containing 15 #g HA would induce protective immunity in horses but multiple doses of vaccine were required. The next question is whether commercial vaccines available in the USA will provide cross-protection against challenge with Eq/Jilin. Single doses of commercially available equine influenza vaccines did protect mice from the lethality of Eq/Jilin and reduced the levels of virus
992
Vaccine, Vol. 11, Issue 10, 1993
replication in the mice (Table 7) but did not provide complete protection. These studies show that there is sufficient antigen in the commercially available vaccines to induce immunity to the H3N8 equine influenza viruses circulating in horses in the USA in 1990 but that there is insufficient antigen to give complete cross-protection against the Eq/Jilin (H3N8) virus.
Standardization of equine influenza vaccines A potential problem with commercially available equine influenza vaccines is that their haemagglutinin content is not standardized. An international study involving 13 laboratories in seven countries 1~ indicates that a highly reproducible method is available for standardizing equine influenza vaccines. The method is based on single radial diffusion and provides a reproducible method of standardizing equine influenza vaccines; it has been used for many years with human influenza vaccines and is the mandated standard adopted internationally. Unfortunately, the necessary vaccine standardization is not mandatory for equine influenza vaccines. It has been established that three doses of aqueous equine influenza vaccine containing 15/~g HA per dose provided protective levels of immunity in horses ~ while adjuvanted vaccines achieve protection with fewer doses 12'13. Correlation between haemagglutinin content in the vaccine and protection in these studies in mice are in agreement with studies in horses, indicating that vaccine standardized in terms of HA content does correlate with protection. Although direct comparison cannot be made between the immune responses in mice and horses to equine influenza vaccines, it is apparent that mice can serve as a model for vaccine evaluation. ACKNOWLEDGEMENTS The authors are grateful to Dr K.K. Brown of Miles Inc. who kindly provided Havlogen C and to Dayna Anderson for manuscript preparation. This work was supported by National Institute of Allergy and Infectious Diseases grants AI-29680 and AI-08831 from the National Institutes of Health, Cancer Center Support (CORE) grant CA-21765 and American Lebanese Syrian Associated Charities. REFERENCES 1 Guo, Y., Wang, M., Kawaoka, Y., German, O., Ito, T., Saito, T. and Webster, R.G. Characterization of a new avian-like influenza A virus from horses in China. Virology 1992, 188, 245 2 Webster, R.G., Hinshaw, V.S., Bean, W.J. Jr, van Wyke, K.L., Geraci, J.R., St Aubin, D.J. and Petursson, G. Characterization of an influenza A virus from seals. Virology 1981, 113, 712 3 Bean, W.J., Sriram, G. and Webster, R.G. Electrophoretic analysis of iodine-labelled influenza virus RNA segment. Anal. Biochem. 1980, 102, 228 4 Kendal, A.P. and Pereira, M.S. Concepts and Procedures for Laboratory-based Influenza Surveillance. US Department of Health and Human Services/Public Health Service, 1982 5 Kida, H., Brown, L.E. and Webster, R.G. Biological activity of monoclonal antibodies to operationally defined antigenic regions of the hemagglutinin molecule of A/Seal/Massachusetts/I/80 (HTNT) influenza virus. Virology 1982, 12~, 38 6 Hartree, E.F, Determination of protein: a modification of the Lowry method that gives a linear photometric response. Anal. Biochem. 1972, 48, 422 7 Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227, 680-685
Equine influenza vaccines: R.G. Webster and T.L. Thomas 8
Oxford, J.S., Corcoran, T. and Mugentooler, A.L. Quantitative analysis of the protein composition of influenza A and B viruses using high resolution SDS polyacrylamide gels. J. Biol. Stand. 1981, 9,483 9 Bean, W.J., Schell, Mo, Katz, J., Kawaoka, Y., Naeve, C., Gorman, O. and Webster, R.G. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J. Virol. 1992, 66, 1129 10 Alstead, A.D., Sahu, S.P., Pedersen, D.D., Saari, D.A., Kawaoka, Y. and Webster, R.G. Pathogenic studies and antigenic and sequence comparisons of A/Equine/Aleska/1/91 (H3N8) influenza virus. J. Vet. Diagn. Invest. (in press) 11 Wood, J.M., Mumford, J.A., Dunleavy, U., Seagroatt, V., Newman, R.W., Thornton, D. and Schild, G.C. Single radial
immunodiffusion potency tests for inactivated equine influenza vaccines. In: Equine Infection Disease V. Proc. Fifth International Conference (Ed. Powell, D.G.) The University Press, Kentucky, 1988, pp. 74-79 12 Wood, J.M., Mumford, J., Folkers, C., Scott, A.M. and Schild, G.C. Studies with inactivated equine influenza vaccine I. Serological responses of ponies to graded doses of vaccine. J. Hyg. (Camb.) 1983, 90, 371 13 Mumford, J., Wood, J.M., Scott, A.M., Folkers, Co and Schild, G.C. Studies with inactivated equine influenza vaccine 2. Protection against experimental infection with influenza virus A/Equine/ Newmarket/79 (H3N8). J. Hyg. (Camb.) 1983, 98, 385
Vaccine, Vol. 11, Issue 10, 1993
993
In 1989 an H3N8 influenza virus caused a severe outbreak of respiratory disease in horses in Jilin and Heilongjiang provinces of Northeastern China 1. Morbidity was 81% and mortality was as high as 20% in some herds. The causative agent has been characterized as a new H3N8 influenza virus that derived its entire genome from an influenza virus in an avian species and the prototype strain is A/Equine/Jilin/1/89 (H3N8) ( Eq/Jilin)X. Since 1989, influenza viruses closely related to the Eq/Jilin strain caused a second outbreak of disease in horses in
Department of Virology and Molecular Biology, St Jude Children's Research Hospital, 332 North Lauderdale, PO Box 318, Memphis, TN 38101, USA. *To whom correspondence should be addressed. (Received 30 June 1992; revised 28 August 1992; accepted 28 August 1992) 0264-410X/93/10/0987--07 © 1993 Butterworth-HeinemannLtd
April 1990 in Heilongjiang province with 48% morbidity and no mortality. No influenza was detected in horses in Northern China or adjacent areas in 1991. The question is whether this virus will appear in horses elsewhere in the world and spread or whether it will behave like the seal influenza virus that caused high mortality for 1 year in seals and then disappeared 2. Since the density of horses is relatively high in adjacent areas of Mongolia, the possibility still exists for the new equine strain to spread. In the present study we determine the efficacy of vaccine to an equine H3N8 strain currently circulating in USA to protect against challenge with the new H3N8 isolate from China. We used a mouse model system for these studies and compared the efficacy of vaccine administered in aqueous suspension and in adjuvant. The studies show that vaccines to A / E q u i n e / K e n t u c k y / 1 2 7 7 /
Vaccine, Vol. 11, Issue 10, 1993
987
Equine influenza vaccines: R.G. Webster and T.L. Thomas
90 (H3N8) (Eq/Kentucky) that are highly efficacious at inducing homologous protection are less efficacious against challenge with Eq/Jilin influenza virus. In contrast, vaccines to Eq/Jilin were efficacious against both Eq/Jilin and Eq/Kentucky. MATERIALS AND METHODS
Viruses A/Equine/Jilin/1/89 (H3N8) (Eq/Jilin) and A/Equine/ Kentucky/1277/90 (H3N8) (Eq/Kentucky) were used in these studies. The viruses were grown in 11-day-old embryonated chicken eggs and were purified by velocity sedimentation through 25-70% sucrose gradients in a Beckman SW28 rotor 3. All studies with Eq/Jilin were done in a biosafety level 3 containment laboratory that was approved for such use by the United States Department of Agriculture.
Virus proteins on polyacrylamide gels were stained with Coomassie brilliant blue and quantified using a Bio Image Visage 110 image analyser (Millipore Corp., Bradford, MA). After standardization by estimation of the HA content the influenza virus was inactivated with 0.025% formalin at 4°C for 3 days. The vaccine was tested for residual infectivity by two blind passages in embryonated chicken eggs and the inactivated vaccine was diluted to give 5.0, 1.0, 0.5 and 0.01 #g HA ml- 1 in phosphate-buffered saline.
Mouse adaptation of equine viruses The Eq/Jilin strain rapidly adapted to mice and was 100% lethal by the third mouse-to-mouse passage 1. The Eq/Kentucky influenza virus was passaged in mice multiple times but the virus caused no mortality. Viruses from the fourth and fifth passages were used in the experiments.
Antisera and specific antibodies
Vaccination and challenge of mice
Antisera were prepared in ferrets to the equine strains as recommended4.
Groups of 12-week-old female Balb/c mice were inoculated intraperitoneally with 1.0 ml of the standardized vaccines. The vaccine was suspended either in phosphate-buffered saline or in 10% Havlogen C in phosphate-buffered saline (an adjuvant used in a commercial equine influenza vaccine and kindly provided by Dr K.K. Brown, Miles Inc.). Four weeks after vaccination, mice were anaesthetized with metophane and inoculated intranasally with 10'~ mouse infectious doses (MID)5o/50 #1 of either mouseadapted Eq/Jilin or Eq/Kentucky. Three days after challenge, four mice per group were killed, and their lungs homogenized and titrated in embryonated chicken eggs for virus infectivity. The remaining mice in each group were observed for 10 days for mortality and were bled for serological studies.
Serological tests HA titrations and haemagglutination-inhibition (HI) tests were performed in microtitre plates with receptordestroying enzyme-treated sera 4. ELISA tests were done as described 5. Pre-immune, 3-week postvaccination, and 10-day postchallenge mouse serum samples were assayed for antibody by each procedure. Neutralization tests were done in embryonated chicken eggs using 100 egg infectious doses (EIDso) of virus.
Preparation of non-standardized inactivated vaccines Purified influenza virus at a concentration of 40000 HA ml-~ was treated with 0.025% formalin at 4°C for 3 days. Mice immunized with high doses of Eq/Jilin and Eq/Kentucky (10000 HA in Freund's adjuvant) were bled and challenged with the respective viruses.
Preparation of standardized inactivated vaccines The protein content of purified influenza virus was estimated using a modification of the Lowry method 6. The haemagglutinin content of the purified virus was determined before formalin treatment using the SDSPAGE system of Laemmli7, modified by Oxford et al. s.
Table 1
RESULTS
Serological relationships among equine influenza viruses Previous studies with chicken antisera established that Eq/Jilin influenza virus is more closely related to avian strains than to human or equine H3 influenza viruses 1. To determine if Eq/Jilin is related to any equine H3 viruses, we tested a wider range of equine strains using ferret antisera in haemagglutination inhibition (HI) assays (Table 1 ). Eq/Jilin was antigenically distinct from
Serological relationships among equine influenza viruses HI antibody titres with ferret antiserum to:
Virus
EQ/Miami/63 F100
EQ/Urg/63 F142
EQ/KY/76 F145
EQ/Font/79 F138
EQ/Alas/91 F101
EQ/Jil/89 F200
EQ/Miami/1/63 EQ/Uruguay/1/63 EQ/KY/1/76 EQ/Fontainebleau/1/79 EQ/KY/1/81 EQ/TN / 5/86 EQ/KY/1277/90 EQ/Alaska/1/91 EQ/Jilin/1/89
160" 40 160 80 160 40 160 < 10 < 10
< 10 40 80 40 20 40 80 10 < 10
< 10 10 320 80 160 160 320 20 40
< 10 < 10 80 40 10 40 80 < 10 < 10
< 10 10 10 20 20 20 10 80 10
< 10-10 10 20 < 10-10 < 10-10 10 10 10 320
aHaemagglutination inhibition titres are expressed as the reciprocal of the highest antiserum dilution inhibiting 4 haemagglutinating units of virus
988
Vaccine, Vol. 11, Issue 10, 1993
Equine influenza vaccines: R.G. Webster and T.L. Thomas
the tested equine influenza viruses isolated between 1963 and 1991. The only detectable cross-reaction was with Eq/Kentucky/76 and this was approximately 100-fold different (320 versus 20; and 320 versus 40). A point of interest is that the Equine/Alaska/1/91 (H3N8) virus is antigenically distinguishable using ferret antiserum from equine influenza viruses isolated from the USA between 1976 and 1990, suggesting that antigenic drift is occurring in equine influenza viruses. These studies with a wider range of equine influenza viruses and ferret antisera confirm the earlier finding 1, obtained with chicken antiserum, that Eq/Jilin is antigenically different from influenza viruses that have circulated in horses since 1963. Will high doses of equine influenza vaccines provide cross-protection against the new Eq/Jilin strain? A crucial question concerning the Eq/Jilin influenza virus is whether vaccines to the H3N8 influenza virus currently circulating in horses in the world will provide protection against the Eq/Jilin strain. Ideally this should be studied in horses but quarantine problems preclude this possibility. Since the Eq/Jilin strain is lethal in mice it provides a satisfactory mouse model system. Eq/Kentucky was adapted to mice and served as a representative of strains currently circulating in the USA. The Eq/Jilin strain grew to high titres in the lungs and killed all the non-vaccinated mice in 10 days; the Eq/Kentucky strain replicated in the mouse lungs but caused no mortality (Table 2). Each of the vaccines induced high levels of HI antibody before challenge but the antibody reacted only with the vaccinating virus; there was no serological cross-reaction between Eq/Kentucky and Eq/Jilin. The vaccines provided complete protection against the homologous strain - no virus was detected in the lungs of the challenged mice and no mice showed signs of disease. In cross-protection studies, neither vaccine provided complete protection but reduced the level of virus replication by three orders of magnitude and both vaccines prevented mortality after Eq/Jilin challenge. HI studies on postchallenge sera showed no detectable cross-reactivity between the Kentucky and Jilin strains. These studies indicate that, despite the absence of detectable serological cross-reactivity in HI tests between the Eq/Jilin and Eq/Kentucky strains, high doses of
vaccine will provide partial cross-protection between these H3N8 viruses. Dose response of equine influenza vaccines administered to mice in aqueous suspension The above studies provide no information on the dose of vaccine required to provide protection against homologous or heterologous challenge, and the vaccines were administered in Freund's complete adjuvant which is unacceptable in horses. Vaccines standardized in terms of haemagglutinin content were administered in aqueous suspension to mice to establish the dose response required to induce protective immunity. Complete protection was taken as the level of vaccine that prevented virus replication in the lungs of mice 3 days after challenge (Tables 3 and 4). Doses of ~>0.1/~g of Eq/Kentucky induced homologous protection; in contrast, Eq/Jilin required doses >~0.5#g for homologous protection (Tables 3 and 4 ). The lowest dose of Eq/Kentucky tested (0.01 #g HA) induced partial homologous protection by reducing the level of virus in the lungs of mice. Cross-protection differed with the vaccine strain; Eq/Kentucky vaccine provided only limited protection against challenge with Eq/Jilin even at the highest doses tested (5/~g HA) (Table 3 ). Although the Eq! Kentucky vaccine did not prevent replication of Eq/Jilin in the lungs at doses from 0.5-5.0 #g HA, it did protect against the lethality of Eq/Jilin (results not shown). On the other hand, Eq/Jilin vaccine at 0.5 #g HA gave complete protection against Eq/Kentucky challenge (Table 4). Serological responses were measured by HI, ELISA and by neutralization of infectivity. The HI assay detected antibody at high doses of antigen (0.5-5.0/~g) but not at low doses of antigen (0,1 #g) (Table 3) and did not correlate with protection. In contrast, ELISA and neutralization assays correlated with protection. Eq/Kentucky vaccine failed to induce cross-reactive antibody whereas high doses of Eq/Jilin vaccines did induce cross-reactive antibodies that were detectable in ELISA and neutralization assays. These studies indicate that equine influenza vaccines ~to the currently circulating strains (Eq/Kentucky) administered in aqueous suspension do not provide complete protection against challenge with the new Eq/Jilin strains. There was reduction in the levels of virus that replicated in the lungs of Eq/Kentucky vaccinated mice after Eq/Jilin challenge.
Table 2 Will high doses of equine influenza vaccines provide cross-protection against the new Equine/Jilin strain? Reactivity in mice
HI antibody to the following viruses:
Vaccine
Challenge virus
Sick/dead/total Day 10 PI
Virus titre (Ioglo) lung day 3 PI
Prechallenge EQ/KY/90
Prechallenge EQ/Jilin/89
Postchallenge EQ/KY/90
Postchallenge EQ/Jilin/89
None EQ/KY EQ/Jilin None EQ/KY EQ/Jilin
EO/Jilin EQ/Jilin EQ/Jilin EQ/KY EQ/KY EQ/KY
6/6/6 0/0/7 0/0/7 0/0/6 0/0/7 0/0/6
>6.5 3.6 (0-6.5) 0 5.75 (5.5-6.3) 0 1.2 (0-4.75)
< 10 328 (0-1280) < 10 < 10 222 (40-1280) <10
< 10 < 10 316 (40-1280) < 10 < 10 225 (40-1280)
Dead 1440 < 10 < 10 5260 <10
Dead < 10 240 < 10 40 160
Groups of ten Balb/c mice were immunized with 10000 HA of formalin-inactivated viruses in Freund's complete adjuvant by the intraperitoneal route. Mice were bled 3 weeks later and were challenged with 10~ mouse infectious doses 4 weeks after initial vaccination. Four mice from each group were killed 3 days postinfection (PI) and assayed for infectious virus. The mice were observed for 10 days for morbidity and mortality and bled 10 days postchallenge
Vaccine, Vol. 11, Issue 10, 1993
989
Equine influenza vaccines: R.G. Webster and T.L. Thomas Table 3
Dose response of Equine/Kentucky and Equine/Jilin vaccines in aqueous suspension in mice
Vaccine: EQ/KY/90 aqueous dose (#g HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Antibody titres to Eq/Kentucky
Challenge virus
Virus titres in mouse lungs day 3 PI
HI
EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL
0 0 0 0 1.9 2.8 4.3 5.0 5.1 7.0 6.8 6.6
8 5 7 < 10 < 10 <10 10 5 3 <10 <10 < 10
Prechallenge
Antibody titres to Eq/Jilin
Postchallenge
ELISA"
Neut.
HI
ELISA"
HI
33 10 7
400 40 <10 ~>560 125 < 10
45 55 45 40 10 <10 90 110 210 13 <10 Dead
100 100 55 33 255 155 55 Dead
<10 <10 < 10 < 10 < 10 <10 <10 <10 <10 <10 <10 < 10
< 25 7 < <
Postchallenge
Prechallenge ELISA = < < < < < < < <
Neut.
HI
< 10 < 10 <10 <10 <10 < 10
<10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10
<
Dead
Dead
ELISA"
< < <
_ _
33 25 7
-, Not tested; PI, postinfection aELISA titres x 10 -3 ( < = < 1000)
Table 4
Dose response of Equine/Kentucky and Equine/Jitin vaccines in aqueous suspension in mice
Vaccine: EQ/Jil/89 aqueous dose (/Jg HA)
Challenge virus
Virus titres in mouse lungs day 3 PI
Antibody titres to Eq/Jilin
5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY
0 0 0 3.1 6.9 6.6 0 0 0 0.8 2.5 2.8
Prechallenge HI
< < <
< <
22 15 12 10 10 10 27 22 18 2.5 10 10
Antibody titres to Eq/Kentucky
Postchaltenge
ELISA"
Neut.
HI
ELISA"
HI
10
~>560 400 < 10 400 40 < 10
20 15 20 20 45 Dead 40 25 25 < 10 5 < 10
33 25 33 Dead 18 25 7 <
<10
< < < 10 7 < <
Postchallenge
Prechallenge ELISA"
Neut.
HI
3
18 18 <10 12 12 <10
<10 <10 <10 <10 <10
<10 <10 <10 <10 < 10 <10 <10 <10 <10 <10 <10
< < < < < < <
ELISA"
10 33 33
Dead
Dead
5 <10 7.5 <10 <10 <10
55 100 55 33
-, Not tested; PI, postinfection aELISA titres x l 0 -3 ( < = <1000)
Dose response to equine influenza vaccines administered in Haviogen C adjuvant The preliminary studies (Table 2) used vaccine in Freund's complete adjuvant to induce partial crossprotection between Eq/Kentucky vaccine and Eq/Jilin challenge. Since Freund's complete adjuvant is unacceptable in horses we tested Havlogen C adjuvant, which has been used in commercial equine influenza vaccines. We determined the dose of antigen in adjuvant required to induce protective immunity and compared it with the studies on vaccine administered in aqueous suspension. Doses of Eq/Kentucky vaccine in adjuvant from 0.01 to 5/~g HA induced homologous protection (Tables 5 and 6). The adjuvant produced at least a tenfold increase in the immunogenicity of the vaccine administered in aqueous suspension. High doses of Eq/Kentucky in adjuvant ( 1.0, 5.0 #g HA ) provided cross-protection from Eq/Jilin challenge and lower doses of Eq/Jilin in adjuvant (0.1-5.0#g HA) provided cross-protection from Eq/Kentucky (Tables 5 and 6). HI antibody titres did not correlate with protection but ELISA titres did correlate both with protection and neutralizing antibody. These studies indicate that administration of equine
990
V a c c i n e , Vol. 11, I s s u e 10, 1993
influenza vaccines in adjuvant increases the immunogenicity of the vaccine both for homologous and heterologous challenge.
Will commercially available equine influenza vaccine protect mice against challenge with equine Jilin virus? Groups of Balb/c mice were immunized with commercial equine influenza virus vaccines available in the USA and were challenged with either Eq/Kentucky or Eq/Jilin (Table 7). Each vaccine provided complete protection against Eq/Kentucky and gave partial protection against Eq/Jilin. The levels of Eq/Jilin in the lungs of vaccinated mice were reduced and the animals were protected from lethal infection. No HI antibody was detected in mice before challenge but ELISA showed detectable levels of antibody. This study shows that commercially available equine influenza vaccines protect against challenge with viruses related to Eq/Kentucky but provided incomplete protection against challenge with Eq/Jilin. The absence of detectable levels of HI antibodies induced before challenge suggests that the dose of HA antigen in the vaccines is in the range 0.01-0.1/~g HA per dose.
Equine influenza vaccines: R.G. Webster and T.L. Thomas Table 5
Dose response of Equine/Kentucky and Equine/Jilin vaccines in adjuvant in mice
Vaccine: EQ/KY/90 adjuvant dose (/~g HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Antibody titres to Eq/Jilin
Antibody titres to Eq/Kentucky
Challenge virus
Virus titres in mouse lungs day 3 P!
HI
ELISA =
Neut.
HI
KY/90 KY/90 KY/90 KY/90 KY/90 KY/90 JIL/89 JIL/89 JIL/89 JIL/89 JIL/89 JIL/89
0 0 0 0 0 2.8 0 0 1.8 2.9 6.0 6.6
32 13 43 30 < 10 < 10 40 30 27 37 < 10 < 10
178 100 70 < 333 178 33 <
>~560 ~>560 >~560 >~560 -
55 120 100 60 20 < 10 240 80 130 120 60 Dead
Postchallenge
Prechallenge
ELISA"
Prechallenge HI
ELISA"
Postchallenge Neut.
HI
ELISA °
-
< 1 0
-
12
< 1 0
-
-
< 1 0
-
-
< 1 0
-
<10 <10 <10 <10
10 7
<10 <10 <10 <10
10 14
333
255 100 33
< <
-
<10
-
-
< 1 0
-
< 10 < 10 <10 < 10
10 3
778
778 100 Dead
<10 -
<10 < 10 -
< <
< <
< 1 0
<10 < 10 < 10 <10 Dead
-
78 55 < Dead
-, Not tested; PI, postinfection =ELISA titres × 10 -3 ( < = < 1000)
Table 6 Vaccine: EQ/Jil/89 adjuvant dose (/zg HA) 5.0 1.0 0.5 0.1 0.01 None 5.0 1.0 0.5 0.1 0.01 None
Dose response of Equine/Kentucky and Equine/Jilin vaccines in adjuvant in mice Antibody titres to EQ/Jilin
Challenge virus
Virus titres in mouse lungs day 3 PI
HI
EQ/JiL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/JIL EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY EQ/KY
0 0 0 0 0 6.6 0 0 0 0 0.8 2.8
80 100 107 60 3 < 10 88 127 107 40 4 < 10
Prechallenge
Antibody titres to EQ/Kentucky
Postchallenge
Prechallenge
ELISA °
Neut.
HI
ELISA"
HI
33 78 10
~>560 -100 >~560 ~>560 -
100 200 267 80 60 Dead 107 150 150 80 120 < 10
78 78 33 Dead 100 25 48 <
< 10 < 10 <10 < 10 < 10 < 10 <10 <10 <10 < 10 < 10 < 10
< 100 25 25 <
ELISA" 33 25 10 < 55 25 25 <
Postchallenge Neut.
HI
ELISA °
40 20 20 60 -
< 10 < 10 <10 < 10 < 10 Dead <10 125 7.5 < 10 < 10 < 10
78 33 33 Dead 255 155 178 33
-, Not tested; PI, postinfection "ELISA titres x 10 -3 ( < = < 1000)
Table 7
Do commercially available equine influenza vaccines protect mice against challenge with Equine/Jilin influenza virus?
Virus titres (Ioglo) in the lungs of mice after challenge with the following viruses: Commercial equine influenza vaccines"
Eq/KY/1277/90
A
<0.5
B C None
<0.5 NT 3.4 (3.25-3.5)
Antibody titres to Eq/Kentucky when challenged with Eq/Kentucky Prechallenge
Postchallenge
Antibody titres to Eq/Jilin when challenged with Eq/Jilin Prechallenge
Postchallenge
Eq/Jilin/1/89
HI
ELISA
HI
ELISA
HI
ELISA
HI
ELISA
4.75(4.25-5.75) 5.1 (4.75-5.5) 4.0 (3.5-4.5) 7.1 (6.75-7.25)
<10 <10 NT <10
1000 1000 NT <1000
20 10 NT <10
100000 100000 100 000 10000
<10 <10 < 10 <10
1000 1000 1000 <1000
<10 80 20 Died
100000 100000 10 000
Groups of six mice were vaccinated intraperitoneally with 0.2 ml of commercially available equine influenza vaccines; 4 weeks after vaccination the mice were challenged with approximately 10" mouse infectious doses of virus in 50 ~1 by the intranasal route. Three days after challenge the mice were killed and the virus titres in the lungs determined in embryonated chicken eggs. Serological tests were done on the remaining mice as given in Materials and methods NT, not tested "These vaccines contain equine 1 (H7N7) and equine 2 (H3N8) viruses but neither the antigen content nor the strains included are revealed
V a c c i n e , Vol. 11, I s s u e 10, 1993
991
Equine influenza vaccines: R.G. Webster and T.L. Thomas
DISCUSSION The appearance of a new equine influenza virus of the H3N8 subtype in horses i that is genetically closely related to avian H3N8 influenza viruses suggests that this particular subtype has the potential to replicate and cause serious disease problems in horses. Phylogenetic analysis of H3 haemagglutinin genes 9 indicates that the H3N8 equine-2 influenza viruses currently circulating in horses also originated from avian species. Thus we might expect that the Eq/Jilin (H3NS) will also have the potential to spread in horses. At this time we do not know if the new H3N8 Eq/Jilin virus will spread to susceptible horses elsewhere in the world. If it does, we need to know the efficacy of available equine vaccines. The appearance of influenza virus in horses in Alaska in 1991 raised the possibility that the Eq/Jilin H3N8 strain had spread; the Alaska/I/91 (H3N8) strain was associated with 100% morbidity and low mortality in some herds 1°. Serological analysis (Table 1 ) and partial sequencing of each gene established that the Alaskan isolate belongs to the equine H3N8 lineage circulating in the USA in 1989-1990. However, it is noteworthy that the Alaskan strain has shown antigenic drift, being at least fourfold different by HI from 1989-1990 equine H3N8 virus in the USA. If the new H3N8 influenza virus spreads outside Northeastern China, will vaccines currently available provide protection? This is the central question addressed in the present study using a mouse model.
Homologous protection These studies indicate that a single dose of aqueous equine vaccine to Eq/Kentucky or Eq/Jilin (0.1/~g of Eq/Kentucky and 0.5/~g of Eq/Jilin) induces complete homologous protection and that tenfold lower doses are efficacious if administered in adjuvant. Serological responses as measured by HI assays did not correlate witl~ protection but ELISA and neutralization assays showed good correlation with protection.
Cross-protection High doses of Eq/Kentucky vaccine administered in aqueous suspension gave reductions in the levels of Eq/Jilin virus in the lungs of mice (Table 3) and the mice did not die. Eq/Jilin vaccine gave significantly more cross-protection against Eq/Kentucky challenge (Table 4). Cross-protection occurred in the absence of detectable cross-reactive HI antibodies but ELISA and neutralization tests correlated with protection. Administration of high doses of either vaccine in adjuvant (1.0/~g HA) provided complete protection against heterologous challenge. No virus was detected in the lungs of infected mice and there was no mortality. Thus high doses of equine influenza virus vaccines administered in adjuvant will provide cross-protection between Eq/Kentucky and Eq/Jilin. Studies in horses 1~ showed that aqueous vaccine containing 15 #g HA would induce protective immunity in horses but multiple doses of vaccine were required. The next question is whether commercial vaccines available in the USA will provide cross-protection against challenge with Eq/Jilin. Single doses of commercially available equine influenza vaccines did protect mice from the lethality of Eq/Jilin and reduced the levels of virus
992
Vaccine, Vol. 11, Issue 10, 1993
replication in the mice (Table 7) but did not provide complete protection. These studies show that there is sufficient antigen in the commercially available vaccines to induce immunity to the H3N8 equine influenza viruses circulating in horses in the USA in 1990 but that there is insufficient antigen to give complete cross-protection against the Eq/Jilin (H3N8) virus.
Standardization of equine influenza vaccines A potential problem with commercially available equine influenza vaccines is that their haemagglutinin content is not standardized. An international study involving 13 laboratories in seven countries 1~ indicates that a highly reproducible method is available for standardizing equine influenza vaccines. The method is based on single radial diffusion and provides a reproducible method of standardizing equine influenza vaccines; it has been used for many years with human influenza vaccines and is the mandated standard adopted internationally. Unfortunately, the necessary vaccine standardization is not mandatory for equine influenza vaccines. It has been established that three doses of aqueous equine influenza vaccine containing 15/~g HA per dose provided protective levels of immunity in horses ~ while adjuvanted vaccines achieve protection with fewer doses 12'13. Correlation between haemagglutinin content in the vaccine and protection in these studies in mice are in agreement with studies in horses, indicating that vaccine standardized in terms of HA content does correlate with protection. Although direct comparison cannot be made between the immune responses in mice and horses to equine influenza vaccines, it is apparent that mice can serve as a model for vaccine evaluation. ACKNOWLEDGEMENTS The authors are grateful to Dr K.K. Brown of Miles Inc. who kindly provided Havlogen C and to Dayna Anderson for manuscript preparation. This work was supported by National Institute of Allergy and Infectious Diseases grants AI-29680 and AI-08831 from the National Institutes of Health, Cancer Center Support (CORE) grant CA-21765 and American Lebanese Syrian Associated Charities. REFERENCES 1 Guo, Y., Wang, M., Kawaoka, Y., German, O., Ito, T., Saito, T. and Webster, R.G. Characterization of a new avian-like influenza A virus from horses in China. Virology 1992, 188, 245 2 Webster, R.G., Hinshaw, V.S., Bean, W.J. Jr, van Wyke, K.L., Geraci, J.R., St Aubin, D.J. and Petursson, G. Characterization of an influenza A virus from seals. Virology 1981, 113, 712 3 Bean, W.J., Sriram, G. and Webster, R.G. Electrophoretic analysis of iodine-labelled influenza virus RNA segment. Anal. Biochem. 1980, 102, 228 4 Kendal, A.P. and Pereira, M.S. Concepts and Procedures for Laboratory-based Influenza Surveillance. US Department of Health and Human Services/Public Health Service, 1982 5 Kida, H., Brown, L.E. and Webster, R.G. Biological activity of monoclonal antibodies to operationally defined antigenic regions of the hemagglutinin molecule of A/Seal/Massachusetts/I/80 (HTNT) influenza virus. Virology 1982, 12~, 38 6 Hartree, E.F, Determination of protein: a modification of the Lowry method that gives a linear photometric response. Anal. Biochem. 1972, 48, 422 7 Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227, 680-685
Equine influenza vaccines: R.G. Webster and T.L. Thomas 8
Oxford, J.S., Corcoran, T. and Mugentooler, A.L. Quantitative analysis of the protein composition of influenza A and B viruses using high resolution SDS polyacrylamide gels. J. Biol. Stand. 1981, 9,483 9 Bean, W.J., Schell, Mo, Katz, J., Kawaoka, Y., Naeve, C., Gorman, O. and Webster, R.G. Evolution of the H3 influenza virus hemagglutinin from human and nonhuman hosts. J. Virol. 1992, 66, 1129 10 Alstead, A.D., Sahu, S.P., Pedersen, D.D., Saari, D.A., Kawaoka, Y. and Webster, R.G. Pathogenic studies and antigenic and sequence comparisons of A/Equine/Aleska/1/91 (H3N8) influenza virus. J. Vet. Diagn. Invest. (in press) 11 Wood, J.M., Mumford, J.A., Dunleavy, U., Seagroatt, V., Newman, R.W., Thornton, D. and Schild, G.C. Single radial
immunodiffusion potency tests for inactivated equine influenza vaccines. In: Equine Infection Disease V. Proc. Fifth International Conference (Ed. Powell, D.G.) The University Press, Kentucky, 1988, pp. 74-79 12 Wood, J.M., Mumford, J., Folkers, C., Scott, A.M. and Schild, G.C. Studies with inactivated equine influenza vaccine I. Serological responses of ponies to graded doses of vaccine. J. Hyg. (Camb.) 1983, 90, 371 13 Mumford, J., Wood, J.M., Scott, A.M., Folkers, Co and Schild, G.C. Studies with inactivated equine influenza vaccine 2. Protection against experimental infection with influenza virus A/Equine/ Newmarket/79 (H3N8). J. Hyg. (Camb.) 1983, 98, 385
Vaccine, Vol. 11, Issue 10, 1993
993