The molecular characterization of influenza virus strains isolated in South Africa during 1993 and 1994

Rex Virol. 1996, 147, 239-245

0 INSTITUT PASTEURIELSEVIER

Paris 1996

The molecular characterization of influenza virus strains isolated in South Africa during 1993 and 1994 T.G. Besselaar

(*), N.K. Blackburn

and B.D. Schoub

National Institute for Virology, Department of Virology, University of the Witwatersrand, Private Bag X4, Sandringham, 2131 (South Africa}

SUMMARY

Influenza A (H3N2) and influenza B viruses isolated recently in South Africa were analysed by partial nucleotide sequencing of the haemagglutinin gene to examine antigenie drift of the isolates relative to the vaccine strains. The genomic analysis of the influenza B isolates revealed a number of differences in the amino acid residues compared with those of the B/Panama/45/90 vaccine strain, and these isolates were found to be antigenically more closely related to B/Quindao/102/91. In both the 1993 and 1994 influenza A (H3N2) isolates significant drii had taken place compared with the H3N2 vaccine strains for those years, emphasizing the need for an influenza vaccine formulated specifically for the southern hemisphere. Key-words: Influenza virus, Vaccine, Haemagglutinin, demiology, Immunogenicity, South Africa.

INTRODUCTION

Influenza virus is a major respiratory pathogen, primarily as a result of the continuous and extensive antigenic variation of the viral surface haemagglutinin (HA). Five major variable sites, designated A-E, have been defined on the globular head region of the HA protein, which is encoded by the HA1 subunit (= 1,064 bases) of the HA gene (Wiley and Skehel, 1987). Monitoring of the antigen@ of the viruses in circulation each year is necessary to identify any new variant strains so that influenza vaccines with antigens closely matching those of the new strains can be prepared annually (Rota et al., 1992). Sequence analysis of the viral HA1 subunit readily allows Submitted

April 4, 1995, accepted February 2, 1996.

I:*) Corresponding author.

Sequencing;

Molecular epi-

the detection of any new mutations which may have occurred relative to previous strains (Rota et al., 1990). The influenza laboratories of the National Institute for Virology currently serve as one of the National Institutes of Influenza of the WHO. An active surveillance programme was established in 1984 to obtain influenza isolates and to monitor influenza epidemics in South Africa (Schoub et al, 1986). The existing programme has recently been extended to include the molecular characterization of local isolates. In this study, influenza A (H3N2) and influenza B viruses isolated during 1993 and 1994 in South Africa were analysed by partial nucleotide sequencing of the HA1 subunit.

240

T.G. BESSELAAR

ET AL.

Table I. Source and passage history of influenza virus isolates. Virus isolate

Source

Passage history

Date of isolation

SA SA SA NIMR NIMR SA SA SA SA SA SA

E2 E2 E2 MK, MDCK MDCK2, E2 El El El El E2 E2

24-6-93 28-6-93 19-7-93 ? 31-1-94 26-5-94 26-7-94 10-g-94 10-8-94 26-8-94 27-9-94

SA SA SA SA SA

E2 E2 E2 E2 E2

18-8-93 25-l l-93 S-6-94 18-7-94 27-7-94

Influenza A (H3N2): A/sAJ643/93 AiSAf658l93 A/S-4/760/93 A/EngIand/276/93 ANalladolidi6/94 AJsAf5 1o/94 AlSAf782l94 A/SAl87Ol94 AlSAl 1194 AlSAl96ll94 AlSAlIO82/94 Influenza

B:

B/SAl909/93 B/SAl1346/93 B/SAl526/94 B/SAf696/94 B/SAJ792/94 E2 = allantoic

fluid

; El = amniotic

MATERIALS

fluid

AND METHODS

Viruses Influenza isolates were mainly obtained through the Witwatersrand Viral Watch Programme as reported by Schoub et al. (1994). In 1993, the majority of influenza isolates were influenza A (H3N2). A total of 22 H3N2 virus isolates were obtained from June-August, while only 2 influenza B strains were isolated in August and one in November. Twenty H3N2 virus isolates were obtained from May-September, 1994, and 30 influenza B viruses were isolated in June-July of the sameyear. The strains selected for sequence analysis were chosen to represent isolates obtained at the beginning, middle and end of each influenza season.The passage history of the influenza A (H3N2) and influenza B virus isolates used in the present study are shown in table I.

HA HI NIMR

= = =

haemagglutinin. haemagglutination National Institute

inhibition. for Medical

PCR amplification of the HA gene RNA was extracted from 100-200 pl of infectious tissue culture fluid, amniotic or ahantoic fluid as outlined by Robertson et al. (1990). The amplification of the HA1 coding region by PCR was carried out using the conditions described by Claas et al. (1992) but with primers specific for the HA gene instead of the NS gene. Briefly, cDNA was synthesized using MMLV reverse transcriptase@-omega) and influenza universal primer (Gubareva et al., 1994) for H3N2 virus and primer B/36/1 (Robertson et al., 1990) for influenza B virus. The HA1 coding region of inlkenza H3N2 cDNA was amplified by PCR using primers H3/4/1 and H3/1077/2 (Robertson and Forsey, 1994). For influenza B cDNA amplification, primers B/36/1 and B/l 140/2 were used (Robertson et al, 1990). The amplification was carried out in 40 cycles in a “Hybaid HB-TRI” cycler (Hybaid Ltd., Middlesex, UK). Each cycle consisted of denaturation at 94°C for 1 min, annealing for 2 min at 55°C for the H3N2 primers and 60°C for the influenza B primers,

PCR SA Research.

= =

polymerase chain reaction South Africa.

INFLUENZA

VIRUS STRAINS

and chain elongation

at 72°C for 2 min. After amplification, 10 ~1 of each reaction was analysed by agarose gel electrophoresis, and the DNA was visualized by ethidium Nucleotide

bromide staining.

sequencing

Prior to sequencing, the amplified products were purified by enzyme digestion using the “SequenasePCR Product Sequencing” kit (United States Biochemical). Partial nucleotide sequencing of the HA1 subunit was performed by direct sequencing of the

purified

DNA using the above kit. For the H3N2

strains, the region of the HA gene chosen for sequencing (500bp) encodes a stretch of 165 amino acids encompassing the receptor-binding site and antigenic regions around it. The oligonucleotide sequencing primers used are shown in table II. The primers are numbered according to the sequence of AlAichii2l68 HA cDNA (Verhoeyen et al., 1980) for the H3N2 HA and according to Rota et al. (1990) for influenza B. Reaction products were separated on 8% polyacrylamide urea sequencing gels, and the data was analysed using the HlBIO DNASIS programme (Hitachi Software Engineering America Ltd., Brisbane, CA, USA).

RESULTS Sequence comparison of the 1993 and 1994 influenza B isolates with the B/Panama/45/90 vaccine strain Common changes between the amino acid sequences of the 1993 South African influenza B

ISOLATED

IN SOUTH

AFRICA

isolates and those of the vaccine strain were seen at residues 73(V-I), 76(T-I), 116(N-H), 122(QH) and 150(D-S) (table Ill). The 1994 isolates were all identical to one another and differed from the B/Panama/45/90 vaccine strain at positions 75 (T-l), 129 (R-K), 15O(D-T), 232 (N-D) and 266(V-I). Sequence comparison of the 1994 B isolates with the B/Quindao/l02/9 1 reference strain revealed that there were differences in six amino acid residues at positions 75(T-I), 146(A-V), 150(S-T), 196(N-D), 198(1-T) and 266(V-I) (table Ill). The presence of aspartic acid (D) rather than asparagine (N) at position 196 correlated with passage in eggs.

Sequence comparison of the 1993 H3N2 lates with the vaccine strain

II.

Primer

Sequencing primers for influenza (H3N2) and influenza B HAl. Nucleotide

Sequence

H3N2: H3/361/1

361-376

H3/669/1

669-683

Y-GCAACTGITACCCITA-3’ 5’-GCATCAGGGAGAGTC-3’

B:

Bl B2 B3 B4 BS

--

36-53 5’-GAAGGCAATAATTGTAC-3’ 256-269 5’-TGCACAGATCTGGA-3’ 370-384 5’-ATGCACGACAGAACA-3’ 601-614 5’-GAAGTACCATACAT-3’ 5’-GGTAGAATTGTTGTI-3’ 805-8 19

iso-

The amino acid sequence changes in the H3 HAS of three South African 1993 influenza isolates relative to the A/Beijing/353/89 vaccine strain are shown in table IV. Changes were seen in 8 to 9 amino acid residues over the area of the HA1 sequenced. Substitutions at residues 145(KN), 156(E-K), 187(A-T), 189(R-S), 190(E-D), 193 (K-S), 262 (T-N) and 276 (T-N) occurred in all three isolates. Of these, the changes at positions 145, 156 and 193 may have been the result of growth in eggs.

Sequence comparison of the 1993/1994 isolates with the vaccine strain Table

241

H3N2

A

Table V shows the differences in the amino acid sequence for representative 1994 H3N2 influenza isolates compared with the A/Beijing/32/92 vaccine strain over the region sequenced. Changes common to both the 1993/94 European isolates and the 1994 South African isolates occurred at residues 145 (S-N), 157 (S-L), 189(R-S) and 276(T-N). The South African isolates exhibited additional differences at residues 216(N-D), 219(S-Y) and 260(1-L). The presence of glutamine (Q) in the vaccine strain and several of the isolates rather than leucine (L) at position 226 correlated with passage in eggs.

242

T.G. BESSELAAK

Table III.

Sequence comparison

ET AL.

of influenza B isolates with B/Panama/45/90 and B/Quindao/lO2/9 1. Substitution at amino acid residue in HA1 (*) 116 122 129 146 150 196 N Q R V D D

B/Panama/45/90

73 V

75 T

76 T

B/Quindao/l02/91

-

-

-

-

-

K

A

S

BlSAi909i93 BJSAfl346l93 B/SA/526/94 BiSAJ696l94 BlSAl792i94

1 I -

I 1 1

1 1 -

H H -

H H -

K K K

_ -

s S T T T

The region corresponding (*’ Amino acid differences by a dash (-).

to IO@1000 bp of the HA1 was sequenced. between the HAls of the isolates and that of the B/Panama/45/90

Table IV. Sequence comparison

Vacc AlBeijl353189 AlSAl643f93 AlSAl658l93 AJSAi76Ol93 The region corresponding (‘1 Amino acid differences

of influenza H3N2

135 E

145 K

K -

N N N

K K K

232 N

266 V

N

I

D

-

-

-

D D D

I I I

vaccine

strain.

1993 isolates with A/Beijing/353/89

T T T

S S S

D D D

No change

is denoted

vaccine strain.

S S S

262 T

276 T

N N N

N N N

to 460-960 bp of the HA1 was sequenced. between the isolates and the vaccine strain.

The amino acid sequence changes between the 1994 South African isolates and the A/Shangdong/9/93 reference strain are shown in table VI.

The 1993/94 European isolates examined in this study (represented by the A/England/276/93 strain in table VI) were very similar to the AlShangdongi9l93 strain. All the South African isolates shared common differences with the reference strain at residues 135 (G-K), 2 14(T-I) and 246(D-N). Viruses isolated later in the season had additional substitutions at positions 216(ND), 219(S-Y), 226(4-L) and 260(1-L). comparison

198 T

Substitution at amino acid residue in HA1 (*) 190 193 156 187 189 E K E A R

Sequence comparison of the 1994 H3N2 isolates with the reference strains A/Shangdong/9/93 and A/Guangdong/25/93

Sequence

vaccine strain

of the South African

isolates with the AlGuangdongl25193 reference strain revealed that common differences occurred at positions 275(A-G) and 299(K-P) (table VI). The changes in the later isolates at positions 216, 219,226 and 260 were the same as those outlined above.

DISCUSSION The effectiveness of influenza vaccines can he greatly impaired unless there is a reasonably good antigenic similarity between vaccine strains and wild viruses (Kendal and Cox, 1985). To date, the strains which were incorporated into vaccines used in South Africa were those recommended by the WHO for the northern hemisphere

INFLUENZA

VIRUS STRAINS

Table V. Sequence comparison

ISOLATED

243

AFRICA

of influenza H3N2 1993194 isolates with AAeijing/32/92

Vacc AfBeijf32192

135 G

145 S

157 S

Substitution 189 R

AfEnglandl276/93 A/ValladolidJ6/94 AlsAl5 1O/94 AfStW381l94 A/SAl961/94 A/SAl1082/94

K K K K K

N N N N N N

L L L L L L

S S S s S S

The region corresponding (*) Amino acid differences

IN SOUTH

vaccine strain.

at amino acid residue in HA1 (*) 214 216 219 226 I N S Q T -

D D D

Y Y Y

260 1

276 T

L L L

N N N N N N

275 G

299 R

L L L -

to 460-960 bp of the HAI was sequenced. between the isolates and the vaccine strain.

Table VI. Sequence comparison of 1994 influenza H3N2 isolates with A/Shangdong/9/93

and AlGuangdong/25/93

strains.

Substitution at amino acid residue in HA1 (*I 216 219 226 246 260 N D I S Q

A’Shangdongl9193

135 G

214 T

A/Guangdong/25/93

K

I

-

-

-

N

-

A

K

A/England/276/93 ABA/s 1o/94 AiSAf782/94 A/SA/870/94 AlSAl881f94 AlSAl961f94 A/SAl1082/94

K K K K K K

I I I I I I

D D D D

Y Y Y Y Y

L L L L L -

N N N N N N N

L L L

_ -

_ -

The region corresponding (*I Amino acid differences

to 460-960 bp of the HA 1 was sequenced. between the HAls of the isolates and that of the A/Shangdong/9/93

winter preceding the winter season in the southem hemisphere. The vaccine formulated for the northern hemisphere, however, does not always correspond to viral strains circulating in South Africa (Schoub et al., 1994). Genomic analysis of influenza viruses in this region will help to influence future international decisions on the vaccine strain formulation best suited for the southern hemisphere. Unlike the situation in Europe and North America, where influenza in the 1993/94 season was caused predominantly by influenza A (H3N2) viruses, influenza B activity was high

strain.

during the 1994 winter season in South Africa. Analysis of the region of the HA1 sequenced (900 bp) for the South African 1994 viruses isolated at the beginning, middle and end of the influenza B season revealed that they were all identical to one another. These isolates did, however, exhibit some drift compared with those circulating in this country in 1993 and with the BlPanama/45/90 vaccine strain. The changes at amino acid residues 75, 129, 150, 232 and 266 relative to the vaccine strain were consistent with those seen in certain of the 1994 isolates from Zambia, Norway and Singapore submitted to the

244

T.G. BESSELAAR

National Institute for Medical Research (Mill Hill, London) for characterization (Dr A. Hay, pers. comm.). These viruses are distinguishable from those previously prevalent in Europe, North America and Southeast Asia and are more closely related to the antigenic variant B/Quindao/102/91. The findings here, that the South African 1994 influenza B viruses have exhibited some drift from the vaccine strain, thus highlight the importance of careful monitoring of influenza B virus strains isolated during the next winter season. With regard to the H3N2 isolates, the results of the present study clearly demonstrate that the viruses which circulated in South Africa in 1993 and 1994 were antigenically significantly different from the recommended H3N2 vaccine strains for these years. The 1993 South African H3N2 isolates shared eight common amino acid differences over the region of the HA1 subunit sequenced compared with the A/Beijing/353189 vaccine strain, with substitutions occurring in the antigenically important sites A, B and C. Similarly, the 1994 H3N2 viruses exhibited considerable drift from the recommended AiBeijing/32/92 vaccine strain, with changes seen in antigenic sites A, B, C and D. Unlike the 1993/1994 European H3N2 isolates which were antigenically closely related to the A/Shangdong/9/93 reference strain, the 1994 South African isolates were more closely related to AIGuangdongl25193. Sequence analysis revealed that the first H3N2 isolate of the 1994 winter season (A/SA/510/94) shared a high degree of homology with the A/Guangdong/25/93 strain, with differences in only two amino acid residues over the region sequenced. The fact that the H3N2 viruses isolated later in the season exhibited additional changes in amino acid residues relative to A/Guangdong/25/93 nevertheless suggests that there is a likelihood of further antigenic drift of H3N2 viruses which may circulate in South Africa during the 1995 winter. The importance of the further drift observed in the South African isolates is highlighted by the recommendation that one of the Johannesburg H3N2 strains (A/Johannesburg/33/94) be included in the influenza vaccine

ET AL.

for the Northern Hemisphere 1995/96 season (Wkly. Epidemiol. Rec., 1995).

winter

Acknowledgements This work was supported by a grant from the Polio Research Foundation.

Caract&isation grippal

molCculaire de souches de virus isol6es en Afrique du Sud en 1993 et 1994

Les virus grippaux A (H3N2) et B rCcemment isolCs en Afrique du Sud ont Ctt analysts par un sCquen$age nucltotidique partiel du g?.ne de l’hkmagglutinine pour 1’ tvaluation des tendances antigeniques des souches isoltes par rapport aux souches vaccinales. L’analyse gCnomique des souches B r&vtle des differences au niveau des

rtsidus

acides aminks,

par comparaison

avec la

souche vaccinale B/Panama/45/90, et ces souches se montrent antigtniquement plus proches de la souche vaccinale B/Quindao/l02/9 1. Chez les virus A (H3N2) isolts en 1993 et en 1994, un profil antigknique significatif est observt par rapport aux souches vaccinales H3N2 de ces armies, ce qui souligne la n&essitC d’un vaccin antigrippal formu spCcifiquement pour l’hCmisph&re sud. Mots-de’s : Virus Grippal, Vaccin, Hkmagglutinine, Sequenqage ; EpidCmiologie mol6culaire. ImmunogCnicitC, Afrique du Sud.

References Class,E.C.J., Sprenger,M.J.W., Kleter, G.E.M., van Beek, R., Quint, W.G.V. & Masurel, N. (1992), Type-specific identificationof influenzavirusesA, B andC by the polymerasechain reaction.J. Viral. Methods, 39, 1-13. Gubareva, L.V., Wood, J.M., Meyer, W.J., Katz, J.M., Robertson,J.S., Major, D. & Webster,R.G. (1994), Codominantmixtures of virusesin referencestrains of influenzavirus dueto hostcell variation. Virology, 199,89-97. Kendal, A.P. & Cox, N.J. (1985), Forecasting the epidemic potential of influenza virus variantsbasedon their molecularproperties.Vaccine, 3, 263-266. Robertson,J.S. & Forsey, T. (1994), Applicationsof PCR to influenza, measlesand mumps,in “The PolymeraseChainReaction(PCR) for HumanViral Diagnosis” (J.P. Clewley) (pp. 81-92). CRC Press,Boca Raton.

INFLUENZA

VIRUS

STRAINS

Robertson, J.S., Bootman, J.S., Nicolson, C., Major, D., Robertson, E.W. & Wood, J.M. (1990), The haemagglutinin of influenza B virus present in clinical material is a single species identical to that of mammalian cell-grown virus. Virology, 179, 35-40. Rota, P.A., Wallis, T.R., Harmon, M.W., Rota, J.S., Kendal, A.P. & Nerome, K. (1990), Cocirculation of two distinct evolutionary lineages of influenza type B virus since 1983. Virology, 175, 59-68. Rota, P.A., Hemphill, M.L., Whistler, T., Regnery, H.L. & Kendal, A.P. (1992), Antigenic and genetic characterization of the haemagglutinins of recent cocirculating strains of infIuenza B virus. J. Gen. Virol., 73, 2737-2742. Schoub, B.D., Johnson, S., McAnerney, J.M., Martin E. & DOS Santos, IL. (1986), Laboratory studies of the

ISOLATED

IN SOUTH

AFRICA

245

1984 influenza epidemic on the Witwatersrand. S. Afr. Med. J., 70, 815-818. Schoub, B.D., Johnson, S., McAnemey, J. & Blackbum, N.K. (1994), Benefits and limitations of the Witwatersrand influenza and acute respiratory infections surveillance programme. S. Afr. Med. J., 84,674-678. Verhoeyen, M., Fang, R., Min Jou, W., Devos, R., Huylebroeck, D., Saman, E. & Fiers, W. (1980), Antigenic drift between the haemagglutinin of the Hong King influenza strains A/Aichii2/68 and ANictoriaf3f75. Nature (Lond.), 286, 77 l-776. Wiley, D.C. & Skehel, J.J. (1987), The structure and function of the haemagglutinin membrane glycoprotein of influenza virus. Ann. Rev. Biochem., 56, 365-394. Wkly. Epidemiol. Rec. (24/2/95) 8, 53-56.