Examination of the envelope glycoprotein of yellow fever vaccine viruses with monoclonal antibodies

Examination of the envelope glycoprotein of yellow fever vaccine viruses with monoclonal antibodies A.D.T. Barrett *§, A. Pryde *~, A.R. Medlen*, T.N. Ledger*, J.E. Whitby*, C.A. Gibson*, M. DeSilva t, D.J. Groves t, D.J. Langley* and P.D. Minor s Two panels of envelope glycoprotein reactive monoclonal antibodies (mAbs) were prepared against yellow fever (YF) 17D vaccine viruses. Five mAbs were prepared against the World Health Organization 17D-204 avian leukosis virus-free secondary seed virus and eight mAbs against 17DD vaccine manufactured in Brazil. The majority of these mAbs were type-specific and displayed differing reactions in neutralization tests. One, B14, would only neutralize YF vaccine virus grown in invertebrate cells. Others would differentiate 17D-204 and 17DD vaccines, from different manufacturers, in neutralization tests when the viruses were grown in vertebrate cells. The data indicate that heterogeneity exists between the epitopes that elicit neutralizing antibody on YF vaccine from different manufacturers.

Keywords:Yellowfever;monoclonalantibodies;flavivirus Introduction Yellow fever (YF) has diminished as a major public health problem because of the development of live attenuated vaccines from two distinct origins 1. The French neurotropic vaccine (FNV) was developed by passage of the wild-type French viscerotropic virus (FVV) through mouse brain 2, while the 17D vaccines were developed by passage of the wild-type Asibi virus through chicken embryo tissue 3. Although the FNV was a successful vaccine, cases of postvaccinal accidents were reported in children under the age of 14 years and this resulted in the vaccine losing popularity and it was discontinued in 1980. In comparison, the 17D-derived vaccines have proved extremely safe and are only contraindicated in infants under the age of 6 months. Two substrains of 17D are recognized, known as 17D-204 and 17DD, which were derived by differing passage series of 17D. The 17D-204 vaccines are between passage 233 and 240 from Asibi while 17DD are between passages 286 and 288 (Ref. 4). Currently some 11 centres around the world manufacture 17D-204 or 17DD vaccine. Recently Buckley and Gould 5 have reported that monoclonal antibodies (mAbs) raised against 17D-204 manufactured in the UK would distinguish vaccine from different manufacturers in neutralization tests. Due to the interest of our laboratory in immunogenicity of YF vaccine viruses we raised two panels of envelope protein reactive mAbs, one against a 17DDepartments of Microbiology* and Biochemistry t, University of Surrey, Guildford, GU2 5XH, UK. *'Division of Viral Products, National Institute of Biological Standards and Control, Blanche Lane, South Mimms, Potters Bar, Herts, EN6 3QG, UK. ~Author to whom correspondence should be addressed. ~Present address: Retrovirus Research Laboratory, Wilkie Building, Department of Surgery, University of Edinburgh Medical School, Teviot Place, Edinburgh, EH8 9AG, UK. (Received 11 August 1988; revised 23 February 1989) 0264-410X/89/040333-04$03.00 © 1989Butterworth & Co. (Publishers) Ltd

204 and the other against a 17DD vaccine virus to investigate heterogeneity between vaccines from different manufacturers.

Materials and methods Viruses The viruses used and their origins are shown in Table 1. Each vaccine virus was obtained as a freeze-dried dose of commercial vaccine. This was reconstituted in deionized distilled water and used to prepare a seed stock by infection of human adenocarcinoma SWl3 or monkey kidney Veto cells. For working stocks, each virus was given one further passage in SW13, Vero or Aedes albopictus larvae C6-36 cells. Plaque assays These were performed in six-well dishes (NUNC) containing monolayers of Vero cells. The medium [Eagle's minimal essential medium (EMEM) (Gibco-BRL)] containing 5% newborn calf serum (Seralab)) was removed, the monolayers washed once with phosphate-buffered saline (PBS) and each well inoculated with 100 #1 of the appropriate dilution of virus. Following incubation for 30 min at room temperature, the wells were overlaid with 6 ml of a mixture of 2% agar (Sigma) and double strength EMEM containing 4% fetal calf serum (FCS) (Seralab) and 0.4% DEAE-Dextran (Sigma). After 6 days incubation at 37°C, 2 ml of a second 'overlay' containing 0.008% neutral red (Gibco-BRL) was added to each well. Following overnight incubation at 37°C in the dark, plaques were recorded via the use of a light box (Genetic Research Instruments).

Plaque reduction neutralization tests These were performed using the plaque assay technique described above except that the virus dilutions were incubated for 45 min at 24°C with a 1/20 dilution ofmAb

Vaccine, Vol. 7, August 1989 333

Examination of yellow fever vaccine virus envelope glycoprotein: A. D. T. Barrett et al. Table 1

Viruses used in the studies

Virus

Origin

Lot No.

17D-204-WHO

World Health Organization avian leukosis virus-free secondary seed Wellcome Biotechnology Ltd, UK Central Research Institute, India Connaught Labs, USA Fundacao Oswaldo Cruz, Brazil Instituto Nacional de Salud, Colombia Institut Pasteur, Dakar, Senegal Institut Pasteur, Dakar, Senegal

YF/1/274 E76 6145 182A 284 81.1 IP*

17D-204-U K 17D-204-1nd 17D-204-USA 17DD-Braz 17DD-Col 17DD-Sen FNV

*Obtained from the Institut Pasteur, Paris, France

containing ascitic fluid prior to infecting the Vero cell monolayer. Neutralization titres are expressed as the loglo reduction in virus titre.

Passive protection studies These were performed as described by Gould et al. 6 except that female mice 4--5 weeks old and weighing 25 g (strain TO; Olac, UK) were used. A minimum of eight mice were used in each experimental group.

Indirect immunofluorescence Cultures of SW13 or C6-36 cells were infected with viruses at a multiplicity of infection of 0.1 and incubated at 37°C for 2-4 days depending on the virus. The cells were trypsinized, washed in PBS and resuspended at 10 4 cells in 25 pl of medium (per spot), were placed on PTFE-coated spot slides (Hendley-Essex) and incubated for 4 h at 37°C in a moist atmosphere. The spot slides were washed twice in PBS, then in ice-cold acetone and finally fixed by placing in ice-cold acetone for 15 min. Slides were then stored at - 20°C prior to use. For immunofluorescence the spot slides were thawed at room temperature and then incubated for 30 rain at 37°C with 25 ~l/spot of mAb, test or positive or negative control samples diluted appropriately, depending on the antibody. The spot slides were washed in PBS, dried and 25 #1 of 1:40 dilution of goat anti-mouse polyvalent immunoglobulin conjugated to fluorescein isothiocyanate (Sigma) added to each well and the slides incubated for a further 30 rain at 37°C. The spot slides were washed in PBS, then distilled water and finally air-dried. Following the addition of mountant (10% saline in glycerol) the cells were examined under ultraviolet light using a Leitz dialux 20 epifluorescent microscope for fluorescence. Results of indirect immunofluorescence tests were obtained by two individuals reading slides independently.

centrifuged at 1800 rev. min -1 for 5 min. One ml 50% polyethylene glycol 1500 (Sigma) in L15 medium (GibcoBRL) was then slowly added to the pellet over 1 min. After the mixture had been swirled at 37°C for 1 rain the cells were resuspended in 10 ml L15 medium over a period of 6 min. The suspension was then centrifuged for 5 min at 1000 rev. rain- 1. The pellet was gently resuspended in 25 ml L15 containing 20% FCS and HAT (10 pM hypoxanthine, 0.04 #M aminopterin and 1.6/ZMthymidine). Cell suspension (75 pl) was placed in each well of fiatbottomed 96-well microtitre plates and a further 200 #1 of the above medium added to each well. The plates were incubated at 37°C and fed every 3-4 days with fresh medium. Colonies of hybridomas appeared 10-14 days after the fusion and the culture supernatants were screened by indirect immunofluorescence for the presence of antibodies to YF virus. Positive wells were selected for cloning and negative wells reincubated and retested for antibody at 20 days postfusion. Any remaining negative wells were rejected. Hybridomas from positive wells were cloned by serial twofold dilution of the cells and clones obtained at the end-point were retested for the secretion ofanti-YF virus antibody. Those clones still positive were expanded, in media containing HAT or HT (i.e. no aminopterin) depending on the hybridoma, and grown as ascitic tumours in pristane (Sigma) primed Balb/c mice.

Protein specificity o f the mAbs This was determined by radioimmune precipitation as described by Gould et al. 7

Isotyping o f mA bs This was initially undertaken by immunodiffusion and the results confirmed for selected mAbs using a mouse monoclonal antibody isotyping kit (Amersham International).

Results

Production of hybridoma cell lines Two panels of hybridoma cell lines were prepared as described in Materials and methods against two YF vaccine viruses. One was 17D-204-WHO (series A hybridomas) and the other was 17DD-Braz (series B hybridomas). Five series A (A1, A2, A3, A4 and A6) and eight series B (B14, B19, B26, B36, B37, B39, B44 and B45) hybridomas were identified. All secreted mAb of the IgG 1 subtype and were shown to recognize epitopes on the envelope glycoprotein by radioimmune precipitation of viral proteins from infected cells (data not shown).

Preparation of hybridoma cell lines secreting mAbs Six-week-old female Balb/c mice were immunized with either 105 p.f.u, of 17DD from Brazil (17DD-Braz) or the World Health Organization 17D-204 avian leukosis virus-free secondary seed virus (17D-204-WHO) by intraperitoneal injection. Three weeks after the initial immunization the mice were boosted with the same dose by the same route. This procedure was repeated twice, four and five weeks after the initial injection. Three or four days after the final boost the spleen was removed from the mouse and teased through a wire mesh to prepare a single cell suspension. Approximately 108 cells were mixed with l073 NS1 myeloma cells and the cell mixture

334

Vaccine, Vol. 7, August 1989

Antigenic reactivities o f mAbs The reactions of the mAbs with a variety of flaviviruses (grown in SW13 cells) in indirect immunofluorescence tests are shown in Table 2. Using a range of flaviviruses, the majority of the mAbs proved to be YF type-specific (AI, A2, A3, A4, B14, B36, B37, B39 and B45) and react with all strains of YF virus examined inclusive of YF vaccines (17DD, 17D-204 and FNV) and wild-type YF virus (Asibi and FVV from Africa and 1899/81 from South America). MAbs A6, B19, B26 and B44 recognized other flaviviruses but did not appear to be flavivirus group common in reaction. Identical results were obtained with

Examination of yellow fever vaccine virus envelope glycoprotein: A. D. T. Barrett et al. Table 2

Reactions of mAbs in indirect irnmunofluorescence tests with a variety of flaviviruses* Reaction with virus

YF mAb

17D-204

17DD

FNV

Asibi

1899/81

FVV

LGT

A1 A2 A3 A4

+ + + +

+ + + +

+ + + +

+ + + +

+ + + +

+ + + +

. . . .

A6

+

+

+

+

+

+

--

B14

+ + + +

+ + + +

+ + + +

+ + + +

+ + + +

+ + + +

.

+

+

+

+

+

+

+

+

+

+

+

+ +

+ +

+ +

+ +

+ +

B19 B26 B36

B37 B39 B44 B45

JE . . . .

. . . .

+ --

.

Den-4

--

--

-+

NT --

+

--

. . . .

+ .

WN

.

+ --

.

.

.

.

+

. .

. .

. .

. .

+ +

.

.

-.

.

*Indirect immunofluorescence tests were performed as described in Materials and methods. 17DD, 17D-204 and FNV refer to the viruses shown in Table 1. Virus abbreviations are those used in the arbovirus catalogue. + , Positive reaction; --, no reaction in indirect immunofluorescence tests; NT, not tested Table :3 Plaque reduction neutralization tests of YF vaccine viruses

from different manufacturers grown in human SW13 cells by mAbs

Table 4

Plaque reduction neutralization tests of 1713-204 and 17DD vaccines grown in mosquito cells by mAbs

Antibody titre with virus

17D-204 mAb A1 A2 A3 A4 A6 B14 B19 B26 B36

USA 0 NT 0 0 0

Ind 0 0 0 0 0

WHO 0 1 0 0 2

Antibody titre

17DD UK 1 NT 0 0 0

Col 0 0 0 0 0

Braz 0 0 0 0 0

mAb Sen 0 0 0 0 0

FNV 0 0 1 0 2

17D-204-WHO

17DD-Braz

A1

0

0

A2 A3 A4 A6

0 0 0 0

0 0 0 0

2

2

0

0

0

0

0

NT

0

0

0

0

B14 B19 B26

0

0

0

0

0

0

B36

0

0

0 0 0

0 NT 0

0 0 0

0 1 0

0 3 0

0 0 1

B37 B39 B44

0 0 0

0 0 0

0 0

4 1

0 0

0 0

0 0

0 0

0 2

B45

0

0

0

0

0

0

0

0

1

Antibody titres are expressed as log10 reduction in virus titre. Experiments were performed as described in Materials and methods

B37

NT 0 0 NT 0

NT NT 0 NT 0

B39 B44

0 0

B45

1

Antibody titres are expressed as log10 reduction in virus titre. Experiments were performed as described in Materials and methods. NT, not tested

the reactivities of mAbs against virus grown in mosquito C6-36 cells.

Analysis o f epitopes on YF vaccine viruses that elicit neutralizing antibodies Preliminary studies showed that the mAb-containing ascitic fluids contained similar amounts of protein (1.9 +0.2 mg ml-1). Therefore plaque reduction neutralization tests were used to investigate the ability of each of the mAbs to neutralize YF virus (Tables 3 and 4). Using F N V - N T grown in SWl3 cells, five of the 13 mAbs (A3, A6, B37, B44 and B45) neutralized the virus, while B36 was the only mAb which would neutralize 17DD vaccines. The latter mAb neutralized 17DD-Braz and 17DD-Sen but failed to neutralize 17DD-Col. Also, B36 was far more effective at neutralizing 17DD-Sen (3 loglo units) than 17DD-Braz (1 loglo unit). 17D-204 vaccine from four centres (USA, India, U K and WHO) were examined and six (A1, A2, A6, B39, B44 and B45) of the 13 mAbs demonstrated neutralization of selected vaccine strains.

These mAbs would each only neutralize vaccine from one of the four centres tested, but the vaccine neutralized depended on the mAb used. For example, mAb A 1 would only neutralize 17D-204-UK while B44 only neutralized 17D-204-WHO (Table 3). Interestingly, mAb B39 neutralized 17D-204-WHO vaccine to high titre (4 loglo units) but failed to neutralize any of the other vaccines tested including 17DD-Braz, the virus it was prepared against. All of the above results were obtained using virus grown in vertebrate SWl3 cells. Since YF virus is an arbovirus, the mAbs were used to neutralize virus grown in mosquito C6-36 cells (Table 4). Of the two viruses examined (17DD-Braz and 17D-204-WHO), only one (B14) of the 13 mAbs would neutralize virus grown in mosquito cells and neutralized both viruses to the same extent.

Passive protection of mice by mAbs Since protective immunity can be analysed in terms of passive immunity as well as neutralizing antibody response, the mAbs were used in passive protection studies (Table 5). None of the 13 mAbs would protect mice against challenge with F N V - N T or 17DD-Braz, while only two

Vaccine, Vol. 7, A u g u s t 1989

335

Examination of yellow fever vaccine virus envelope glycoprotein: A. D. T. Barrett et al. Table 5

Passive protection of mice with YF vaccine viruses by mAbs Mice surviving (%) challenged with

mAb

17D-204-WHO

17DD-Braz

17DD-Col

FNV-NT

A1 A2 A3 A4 A6

67 62.5 55 37.5 55

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

B14 B19 B26 B36 B37 B39 B44 B45

0 0 0 0 37.5 44 67 70

0 0 0 0 0 0 0 O

0 14 0 0 0 0 0 20

0 0 0 0 0 0 0 0

C7a

0

0

0

0

Results are expressed as the percentage of mice which survived intracerebral challenge with virus after pretreatment with mAb. Experiments were performed as described in Materials and methods. aC7 was a control mAb of IgG1 subtype which does not recognize YF virus

mAbs (B19 and B45) protected mice against challenge with 17DD-Col. However, the latter two mAbs only protected 14-20%. In contrast, nine of the 13 mAbs protected mice to some extent against challenge with 17D-204-WHO. This ranged from 37.5% with mAb A4 to 70% with mAb B45.

Discussion The preparation of two panels of hybridomas, one prepared against 17DD-Braz and the other against 17D-204-WHO, is reported. Of the 13 envelope glycoprotein reactive mAbs prepared, eight were against the former and five against the latter virus. The majority of the mAbs were YF type-specific and the remainder were flavivirus intermediate in specificity. The plaque reduction neutralization tests show that only one mAb, B14, was able to neutralize virus grown in mosquito A. albopictus C6-36 cells whilst many of the mAbs neutralized virus grown in vertebrate SWl3 cells. This result could not simply be explained on the basis of lack of the epitopes on virus grown in mosquito cells, as indirect immunofluorescence tests showed that the epitopes recognized by the mAbs were present on viruses grown in both cell types. Also, the results could not be explained on the basis of different concentrations of antibody in the ascites as examination of the ascites showed that they contained similar amounts of protein (1.9 + 0.2 mg ml-1). Grady and Kinch ~ have reported a similar phenomenon with a bunyavirus and we have also observed hostdependent neutralization with Japanese encephalitis virus (Wills et al., in preparation). Presumably such hostdependent neutralization relates to presentation of epitopes on the surface of the virus. This seems likely as it is known that the envelope protein is glycosylated whilst mosquito cells lack sialic acid and the phospholipid content of invertebrate cells differs from that of vertebrate cells 9. This area deserves more research as it is conceivable that preparation of vaccines in invertebrate cells may not result in the expression of epitopes involved in protective immunity in vertebrates. It was also found that epitopes eliciting protective immunity (i.e. passive protection and neutralizing antibody) varied between YF vaccines that had been grown in

336

Vaccine, Vol. 7, August 1989

human SW13 cells. In particular, the 17D-204 vaccines examined, produced from four different centres, exhibited differing epitopes that elicited neutralizing antibody indicating heterogeneity between the same vaccine manufactured in different parts of the world. Since Buckley and Gould 5 and Gould et aL e have reported similar results using mAbs prepared against 17D-204 vaccine manufactured in the UK, it would appear that variation in epitopes that elicit neutralizing antibody and passive protection is a common phenomenon of YF vaccines manufactured in different centres around the world. Why this takes place is not clear but may relate to different methods of vaccine production, including the source of eggs used to prepare the vaccine and growth conditions of the vaccine virus 1. Whether or not these differences are of practical consideration is uncertain. First, no comparison has been made on the efficacy of vaccines produced by different manufacturers and second, the neutralization differences have been detected using monoclonal antibodies raised in mice and the extrapolation to the human situation must be questioned. Although YF vaccines may be very safe and effective, there is clearly heterogeneity in the vaccine manufactured in different centres 1 and further studies are required to investigate whether or not this is of any practical consequence. In particular, analysis of the nucleotide and amino acid sequence of the envelope protein gene of vaccine viruses from different manufacturers may give an insight into the different neutralization results described above.

Acknowledgements A.P. was supported by a SERC-CASE studentship, T.N.L. is supported by a SERC studentship, C.A.G. by the Medical Research Council and D.J.G. by the Agricultural and Food Research Council. We thank Dr F. Rodhain (Institut Pasteur, Paris, France) and the yellow fever vaccine manufacturers for donating samples of their vaccines for use in these studies. We also thank Janet Bunker, Lee Dunster and Alan Jennings for advice with the manuscript.

References 1 Barrett, A.D.T. Yellow fever vaccines. Bull. Inst. Pasteur 1987, 85, 103 2 Mathis, C., Sellards, A.W. and Laigret, J. Sensibilit6 du Macaca rhesus au virus de la fi6vre jaune. C.R. Aead. Sci. (Paris) 1928,188, 604 3 Theiler, M. and Smith, H.H. The use of yellow fever virus modified by in vitro cultivation for human immunization. J. Exp. Med. 1937, 85, 787 4 Monath, T.P., Kinney, R.M., Schlesinger, J.J., Brandriss, M.W. and Bres, P. Ontogeny of yellow fever 17D vaccine: RNA oligonucleotide fingerprint and monoclonal antibody analyses of vaccines produced world-wide. J. Gen. Virol. 1983, 64, 627 5 Buckley, A. and Gould, E.A. Neutralisation of yellow fever virus studied using monoclonal and polyclonal antibodies, J. Gen. Virol. 1985, 66, 2523 6 Gould, E.A., Buckley, A., Barrett, A.D.T and Cammack, N. Neutralising (54k) and non-neutralising (54k and 48k) monoclonal antibodies against structural and non-structural yellow fever virus proteins confers immunity in mice, J. Gen. Virol. 1986, 67, 591 7 Gould, E.A., Buckley, A., Cammack, N., Barrett, A.D.T., Clegg, J.C.S., Ishak, R. and Varma, MG.R. Examination of the immunological relationships between flaviviruses using yellow fever virus monoclonal antibodies. J. Gen. Virol. 1985, 66, 1369 8 Grady, L.J. and Kinch, W. Two monoclonal antibodies against La Crosse virus show host-dependent neutralising activity. J. Gen. Virol. 1985, 66, 2773 9 Stollar, V., Stollar, B.D., Koo, R., Harrap, K.A. and Schlesinger, R.W. Sialic acid contents of Sindbis virus from vertebrate and mosquito cells. Equivalence of biological and immunological properties. Virology 1976, 69, 104