A comparative study of attenuated influenza viruses

Bull. Org. mond. Sante 1971, 44, 593-598Bull. Wid Hith Org. A Comparative Study of Attenuated Influenza Viruses A. S. BEARE,1 H. F. MAASSAB,2 D. A. J. TYRRELL,3 A. N. SLEPUKIN4 & T. S. HALL1 Influenza A and influenza B viruses were adapted to growth at 25°C. When given to volunteer subjects, the viruses were attenuated but remained infective and antigenic. The minimum immunizing dose ofan egg-adapted virus appeared to be 105-0 EID5o. Cloning by plaque selection at 25°C gave seed cultures of relatively low infectivity titres. These titres were increased when necessary by passage at 33°C. No reversion to virulence was observed. Viruses attenuated in the United Kingdom and the USA were compared in volunteer trials with vaccine strains that had already been used in the USSR for mass immunization. Results were broadly similar. Currently available methods of attenuation and work with temperature-sensitive mutants are reviewed. Workers in the USSR originally suggested that influenza A viruses resistant to serum inhibitors were attenuated for man and that they could be used for mass vaccination (Solov'ev et al., 1961, 1968; Smorodincev et al., 1965, 1967). This has since been confirmed in the United Kingdom where volunteer trials showed that inhibitor-resistant strains were less virulent than the inhibitor-sensitive stocks from which they were derived and were potentially immunogenic and protective (Beare, 1969; Beare & Bynoe, 1969). The trials in the United Kingdom also involved a preliminary study of virus attenuation by passage at 25°C. The latter method has received greater attention in the USA (Maassab, 1967, 1968, 1969) where it was shown that influenza A and influenza B viruses, initially virulent for ferrets and mice, could be attenuated for these animals. In this article the responses of volunteer subjects to viruses manipulated in diffe- rent laboratories by a variety of techniques are compared and an attempt is made to assess these viruses as potential human vaccines. 1 Medical Research Council Common Cold Unit, Harvard Hospital, Salisbury, England. 'Michigan University School of Public Health, Ann Arbor, Mich., USA. 'Clinical Research Centre, Watford Road, Harrow, Middlesex, England. ' WHO Research Fellow at the Common Cold Unit. Permanent address: D. I. Ivanovsky Institute of Virology, Academy of Medical Sciences, Moscow, USSR. MATERIALS AND METHODS Viruses The adaptation of A2/Hong Kong/1/68 to growth at 25°C was carried out at the Common Cold Unit as previously described (Beare & Bynoe, 1969), the virus being passaged 17 times in embryonated eggs at that temperature with 3-day incubation periods. Infectivity titres were generally low and a final passage was therefore made at limit dilution at 33°C. This pool, which had a titre of 108.5 50% egg-infecting doses (EID50) per ml of allantoic fluid, was used in a volunteer trial. The A2/Aichi/2/68 strain of the Hong Kong virus was attenuated at the Michigan University School of Public Health. Monolayers of chicken kidney in plastic Petri dishes were infected and overlaid by methods described elsewhere (Maassab, 1968). After incubation at 25'C, a plaque was selected from which the virus was propagated in series in embryonated eggs at the same temperature (Maassab et al., 1969). In the later passages the incubation periods were 5-6 days. The virus grew reasonably well with moderate titres of haemaggluti- nin and the plaquing system appeared to have hastened the adaptation process. Virus that had received 8 such passages showed a titre of 1080- EID50 in eggs incubated either at 33°C or at 25°C. This virus was given to volunteer subjects in two trials. The virus was then passaged again at 33°C 2672 593 - 594 A. S. BEARE AND OTHERS at a dilution of 10-4 ° and new seed was obtained weeks after the trials. Serological and virus iso- with an infectivity titre of 107-3 EID50 per ml. This lation techniques have been described previously seed was used in a third trial. (Beare et al., 1968). Because of the relative insensi- Virus strain B/AA/1/66 was attenuated at the tivity of monkey-kidney tissue-culture for the isola- Michigan University School of Public Health by tion of Hong Kong virus, the allantoic cavity of the technique used for strain A2/Aichi/2/68. A embryonated eggs was sometimes used instead. plaque selected at 25'C was passaged 10 times in eggs at this temperature. The pool given to volunteer subjects had a titre of 10.-5 EID50 RESULTS per ml. Three viruses from the USSR were also given Influenza viruses had previously been attenuated to volunteers. They were not attenuated by us and for animals by the methods used for the attenuation were supplied as vaccine strains by the D. I. Ivanov- of the viruses at the Michigan University School sky Institute of Virology of the USSR Academy of Public Health (Maassab, 1969), but since these of Medical Sciences. Their names and passage had not been inoculated into man, the Aichi variant histories are as follows: was first given in 3 different dilutions. Groups of (1) A2/Istra/10/69, an inhibitor-resistant Hong volunteers received 1.0-ml amounts of virus diluted Kong strain, received 8 egg passages (3 of them in 10-4.0, 10-3.0, and 10-2.0, and these contained doses the presence of horse serum), 4 passages in chick- of 102.0, 103.0, and 104-0 EID50, respectively. The embryo tissue-culture, and 2 further egg passages. results are shown in Table 1. Although a few The temperature of incubation was 33°-35°C. infections occurred and there were some clinical reactions, it was clear that even the highest virus (2) A2/Hong Kong/1/68, an inhibitor-sensitive dose had failed to infect most of the susceptible vaccine strain of the prototype virus, was prepared volunteers. This could have been due to over- independently in Moscow. A laboratory virus attenuation or to insufficient dosage. Since the that had initially received 2 passages in monkey- seed had already been diluted 10-2.0, it was decided kidney tissue-cultures and 3 in eggs was passaged to prepare a new pool with a higher infectivity a further 21 times in eggs at 35°C. titre. Passage at 25°C might have achieved this, (3) B/Dushanbe/1/66 received 22 egg passages at but as it might also have resulted in over-attenuation, 350C, 2 being made in the presence of horse serum. the next passage was made in 6 leucosis-free eggs at a dilution of 10-4.0 at an incubation temperature Volunteer subjects of 33°C. The new pool had a titre of 107 EID50 per ml, which when diluted 1: 10 and given to 11 Isolation and inoculation of volunteers and more volunteers proved adequately infective and methods used in the grading of clinical reactions antigenic (Tables 1 and 2). Only 1 susceptible have already been described (Tyrrell, 1963). Diluted volunteer now failed to become infected. A detailed viruses were given in 1.0-ml amounts intranasally examination of the responses showed antibody rises by dropper or in 0.5-ml amounts with a coarse comparable with those obtained with viruses atte- hand-sprayer. An arbitrary scoring system, based nuated by other means (Beare & Bynoe, 1969) on the incidence of pyrexia, coryza, and subjective and no apparent increase in virulence. The responses symptoms and on handkerchief counts on successive of volunteers to the Salisbury " cold " virus (Table 1) days following virus inoculation, was also used to showed that a 1: 1000 dilution, which provided an compare the severity of reactions induced by the individual dose of 105 5 EID50, was also infective different viruses. and antigenic and, like the Aichi variant, relatively non-reactive. Lastly, in this trial series, the atte- Assessment of the trials nuated B/AA/1/66 strain was given to 19 volunteers.Most of the susceptibles were infected by doses Criteria were generally the incidence and severity ranging from 105-0 to 105-5 EID50 and clinical of clinical reactions, the presence of virus in nasal symptoms were trivial. Nevertheless, the titre of the washings taken on the 2nd, 3rd, and 4th days seed (106.5 EID50 per ml) was probably low for after inoculation, and rises in haemagglutination- clinical use and needed to be elevated, if possible, inhibiting (HI) antibody titres in the blood 2-3 by further passage at 33°C. COMPARATIVE STUDY OF ATTENUATED INFLUENZA VIRUSES 595 Table 1. Responses of volunteer subjects to inoculation with influenza A and influenza B viruses adapted to growth at 24°C at the Common Cold Unit, Salisbury and at the Michigan University School of Public Health, Ann Arbor Serum haemagglutination-inhibiting Dilutin .Virus Significant antibody rises:Virus strain Dilution dose clinicalof seed (ElDso in 1 ml) reactions a, b in those with inallvolunteers initial titres .24 inalvuter 10-4.0 102.0 1/7 1/6 1/7 A2/AICHI/2/68 10-3.0 103.0 2/7 2/3 2/7 10-2.0 104.0 3/21 3/12 4/20 c A2/Aichi/2/68, after passage at 33°C 10-1.0 106.3 6/11 6/7 8/11 A2/Hong Kong/1i/68 (Salisbury) 10-3.0 105.5 3/9 4/4 4/9 B/AA/1/66 (Ann Arbor) 10-1.0 105.0-105.5 2/19 10/15 10/19 a Number of responses/number of specimens tested. b All reactions shown were graded X mild -. c One serum specimen was lost in this series. Table 2. Responses of volunteer subjects to inoculations immunogenic. However, since they have been used with strain A2/Aichi/2/68 virus * in large-scale field trials in the USSR, they appear to merit further study. Strain A2/Istra/10/69 (the VolunterGraing o clini Reciprocal haemagglu- inhibitor-resistant strain) seemed less antigenic than Voute Grading of clinical tination-inhibitingreactions antibody titres before A2/Hong Kong/1/68 but the difference was not and after the trial statistically significant; strain B/Dushanbe/l/66 pro- duced antibody rises in 5 out of 6 initially susceptible 1 Mild <6; 72 volunteers. There were no reports of undesirable 2 Mild <6; 84 clinical effects during mass vaccinations in the USSR, 3 Nil <6; 12 but in the closely supervised trials of the CommonCold Unit a number of reactions were recorded,4 Mild 48; 192 and although most were " mild ", those following 5 Nil 48; 192 inoculation with B/Dushanbe/1/66 were graded 6 Nil 144; 192 " moderate ". 7 Nil 6; 48 An attempt to make a closer comparison of the clinical responses induced by the viruses is shown8 Nil <6; <6 in Table 4. Each of the gradings " mild ", " mode- 9 Mild <6; 24 rate ", and "severe " covered a wide range of 10 Nil 36; 36 symptoms, and initial efforts to give numerical 11 Mild <6; 48 scores to individual reactions suggested that values below 20 signified mild (and therefore acceptable) responses. While a high degree of accuracy is not *Virus dose: 106*3EID5o.The virus had had 8 passages at 25C claimed for the system, this additional analysis and 1 at 33°C. Significant antibody rises are shown in bold-face type. appears to provide useful information. Strain B/Dushanbe/1/66 was clearly under-attenuated, while strain A2/Aichi/2/68 and the Salisbury and Moscow strains of A2/Hong Kong/1/68 were of roughly The responses of volunteers to inoculation with similar pathogenicity and were all adequately atte- the USSR vaccine strains are shown in Table 3. nuated. Strains B/AA/1/66 and A2/Istra/10/69 Both the influenza A viruses were given in doses seemed to produce virtually no reactions in the of 103.0 and 105.0 EIDse, neither of which was fully volunteers infected with them. 596 A. S. BEARE AND OTHERS Table 3. Responses of volunteer subjects to inoculation with live influenza vaccines prepared in the USSR * Incubation Sensitivity EsiaeICiia Serum haemagglutination- Virus strain temperature to horse virus dose Clinical Virus inhibiting antibody rises:during serum ,,. reactions a isolations b attenuation (0C) inhibitor (EID50l in those with b allinitial titres <24 b volunteersb A2/lstra/1 0/69 33-35 Resistant 103.0 0/6 2/6 1/2 2/6 105.0 3/8 2/8 1/3 2/8 A2/Hong Kong/i /68 35 Sensitive 103.0 0/5 2/5 2/3 2/5 105.0 6/9 6/9 4/7 5/9 B/Dushanbe/1/66 35 105.0 3/9 c 5/9 5/6 6/9 * At the D. I. lvanovsky Institute of Virology, Moscow. a Number of responses/number of specimens tested. b These three reactions were graded - moderate -. All others in this table were graded mild . Table 4. Comparison of clinical responses in volunteer subjects infected by the attenuated viruses Meanpeak Mean no. Mean score of Virus strain (ElDus) infected subjectsao handkerchief of reactions per symptoms and(EID5) inectedsubjctsa count infection signs bper infectionb A2/Aichi/2/68 (Ann Arbor) 1 020-1 04.0 7 6.9 0.43 13.3 A2/Aichi/2/68 (Ann Arbor) 106.3 8 11.1 0.6 14.4 B/AA/1/66 (Ann Arbor) 105*°-105.5 11 3.2 0.18 4.8 A2/Hong Kong/1/68 (Salisbury) 105-5 4 21.3 0.75 14.8 A2/lstra/10/69 (Moscow) 103.0-105.0 4 6.3 0.25 6.3 A2/Hong Kong/1/68 (Moscow) 103.0-105.0 8 7.8 0.63 11.8 B/Dushanbe/1/66 105.0 6 21.2 0.6 32.8 a Volunteers with virus excretion or antibody rise or both. b Based on number of days on which there was pyrexia, coryza, increased use of handkerchiefs, and subjective symptoms. DISCUSSION During the attenuation of influenza viruses the influenza-like symptoms which they evoke are The experiments show that influenza viruses replaced by those of afebrile coryza. All the vaccine passaged repeatedly at 25°C, or propagated from strains that we have so far produced cause symptoms plaques selected at this temperature, are relatively of this kind in at least a proportion of the people attenuated for man. Although none of the viruses who are successfully immunized. obtained by these means was of a sufficiently high It was interesting to compare the clinical effects of titre to be used for human immunization, titres were the vaccines from the USSR with those of vaccines subsequently increased by passage at 330C. There from the United Kingdom and the USA. The was then no apparent reversion to virulence. How- degree of reaction with most of the influenza A ever, this latent possibility stresses the desirability, viruses seems to have been about the same, and where possible, of careful initial cloning. since strains from the USSR had been used for COMPARATIVE STUDY OF ATTENUATED INFLUENZA VIRUSES 597 large-scale immunization it is probably an accept- to grow at the upper temperature limits of the able level. Strain B/Dushanbe/1/66 may, however, parent strains and are produced by the use of have been under-attenuated. The attenuation chemical mutagens. Mills and his colleagues demon- methods employed in the USSR reflect the views held strated that two such mutants were attenuated for in that country on the attenuating effect on influenza hamsters and attributed this to the existence of a viruses of simple egg-passage at ordinary tempe- temperature gradient within the respiratory tract, ratures of growth (Smorodincev, 1969; 2danov, which confined the replication of the viruses to the 1967). It has, however, not been possible to atte- upper part of the tract. While we accept the evidence nuate viruses by these means at the Common Cold that ts mutants are attenuated, this explanation seems Unit (Beare et al., 1968; Beare, unpublished data), much over-simplified. Mackenzie's work showed where influenza viruses are held to be relatively that ts defects in WSN virus were associated with stable genetically. The attenuation of strain diminished virulence for mice and that the mutants A2/IstraIlO/69 may well be due to its inhibitor- were infective and antigenic. He was unable to cor- resistant character since such strains are habitually relate lack of virulence with physiological changes in attenuated (Beare & Bynoe, 1969; Slepugkin et al., the haemagglutinin and neuraminidase and con- 1968) and, in our opinion, are a potentially valuable firmed the opinion of others that virulence in in- source of live vaccines providing they do not become fluenza viruses is polygenic. In addition, he thought over-attenuated. For the moment, it is considered that there was no direct connexion between the level that low-temperature adaptation is the most gene- of " cut-off" temperature and the degree of attenu- rally useful of the attenuation methods since it can ation but, in this respect, his views differ from those be applied both to influenza A and influenza B of Mills et al. (1969b), which were also based on serotypes and can, if necessary, be made indepen- Mackenzie's own results. The production of ts dent of plaquing ability. mutants seems unlikely at present to replace low- temperature adaptation as a practical means of In conclusion, reference may be made to publi- obtaining attenuated viruses for live vaccines. cations from other laboratories on the possibility Nevertheless, it provides a system for the genetic of using temperature-sensitive (ts) conditional lethal analysis of influenza viruses, which must ultimately mutants as live influenza vaccine strains (Mackenzie, prove of great importance to the development of 1969, 1970; Mills et al., 1969a). These mutants fail live vaccines ACKNOWLEDGEMENTS We are much indebted to Miss M. D. Turnbull, sponsorship of the Commission on Influenza, Armed S.R.N. and Miss J. Bailey, S.R.N. for help with the Forces Epidemiological Board, and was supported by the surveillance of the volunteers, and to the volunteers for US Army Medical Research and Development Com- their willing co-operation. Mrs P. D. Bradburne, Mrs mand, Department of the Army, under research contract P. K. Brown, and Mrs B. Head provided invaluable No. DADA 17-70-C-0050. technical assistance. TheCommon Cold Unit at Harvard Hospital, Salisbury, The investigation in the USA was conducted under the receives a grant from the World Health Organization. RtSUM12 tTUDE COMPARATIVE DE VIRUS GRIPPAUX?ATTtNUtS Des virus grippaux A et B, attenues par differents pro- l'excretion du virus et la production d'anticorps; b) la cedes dans des laboratoires du Royaume-Uni, des Etats- frequence et la gravit6 des reactions cliniques chez les Unis d'Amerique et d'URSS, ont ete experimentes sur des sujets vaccines; c) l'infectivite des diverses souches virales volontaires a la Common Cold Unit, de Salisbury et la possibilite de les utiliser pour la vaccination de masse. (Angleterre). On se proposait d'etudier: a) les modalites En Angleterre, par passages repet6s sur ceuf, a 25°C, de l'infection et de l'immunisation eventuelle, d'apres suivis d'un dernier passage a 33°C, on a obtenu une 2 598 A. S. BEARE AND OTHERS souche attenuee de virus A2/Hong Kong/1/68 titrant 108,5 DIE58 (dose infectant 50% des embryons de poulet) par millilitre. Aux Etats-Unis, la souche A2/Aichi/2/68 du virus de Hong Kong et une souche de virus grippal B(B/AA/1/66) ont ete attenuees par inoculation au rein de poulet en couche monocellulaire et incubation a 25°C pendant 5-6 jours. Apres repiquage sur ceuf, on a recolte un virus A titrant 106,0 DIE50/ml et un virus B titrant 10WA DIE5o/ml. Trois virus ont et6 attenues en URSS: la souche A2/Istra/10/69 (virus de Hong Kong resistant aux inhibiteurs); une souche A2/Hong Kong/l/ 68 attenu&e par 21 passages sur ceuf, 'a 35°C; une souche B/Dushanbe/l/66 attenuee par 22 passages sur ceuf, a 350C. Les essais sur volontaires ont montre que les differents virus, 'a l'exception peut-etre du virus B/Dushanbe/1/66, etaient convenablement attenues. La technique de la selection de plages permet d'accelerer le processus d'att& nuation, mais les virus ainsi obtenus ne font preuve que d'une infectivite relativement faible; cependant, le pas- sage A 33°C accroit les titres sans augmentation conco- mitante de la virulence. Si l'on en juge d'apres les effets cliniques, I'attenuation des virus obtenus au Royaume-Uni et aux Etats-Unis est du meme ordre que celle des virus d'URSS, dej'a utilises pour la vaccination de masse. Selon les auteurs, I'adaptation des virus a la croissance aux basses temperatures (25°C) apparait comme le pro- cede le plus interessant parmi les methodes d'attenuation dont on dispose actuellement. Elle peut en effet etre appliquee aux s6rotypes A et B et elle ne depend pas necessairement des possibilites de selection de plages. Quant a la production de mutants sensibles a la tempd- rature, bien qu'offrant de l'interet pour l'etude genetique des virus grippaux, elle ne semble pas pouvoir etre substi- tu6e a l'adaptation aux basses temperatures, car elle entraine g6neralement une diminution nette de la virulence. REFERENCES Beare, A. S. (1969) Bull. Wld Hlth Org., 41, 595 Beare, A. S. & Bynoe, M. L. (1969) Brit. med. J., 4, 198 Beare, A. S., Bynoe, M. L. & Tyrrell, D. A. J. (1968) Brit. med. J., 4, 482 Maassab, H. F. (1967) Nature (Lond.), 213, 612 Maassab, H. F. (1968) Nature (Lond.), 219, 645 Maassab, H. F. (1969) J. Immunol., 102, 728 Maassab, H. F. et al. (1969) Bull. Wid Hlth Org., 41, 589 Mackenzie, J. S. (1969) Brit. med. J., 3, 757 Mackenzie, J. S. (1970) J. gen. Virol., 6, 63 Mills, J. et al. (1969a) Bull. Wld Hlth Org., 41, 599 Mills, J. et al. (1969b) Brit. med. J., 4, 690 Slepu.kin, A. N. et al. (1968) Vop. Virus., 13, 224 Smorodincev, A. A. (1969) Bull. Wld Hlth Org., 41, 585 Smorodincev, A. A. et al. (1965) Experiences in the de- velopment of live vaccines against influenza and in- fluenza-like respiratory infections. In: Saunders, M. E., Lennette, E. H., ed., First Annual Symposium on Applied Virology, Boca Raton, Florida 1964, Sheboygan, Wis., Ellis Smorodincev, A. A. et al. (1967) J. Hyg. Epidem. (Praha), 11, 265 Solov'ev, V. D. et al. (1968) Vop. Virus., 13, 396 Solov'ev, V. D. et al. (1961) Vop. Virus., 6, 684 Tyrrell, D. A. J. (1963) Amer. Rev. resp. Dis., 88, 128 Zdanov, V. M. (1967) Present status oflive influenza virus vaccines. In: Proceedings of First International Con- ference on Vaccines against Viral and Rickettsial Diseases of Man, Washington, D.C., Pan American Health Organization, p. 9