Abstract
Five erythrocyte species (horse, goose, chicken, guinea pig, and human) were used to agglutinate avian influenza H5N1 viruses by hemagglutination assay and to detect specific antibody by hemagglutination inhibition test. We found that goose erythrocytes confer a greater advantage over other erythrocyte species in both assays.
Endemicity of H5N1 avian influenza in Southeast Asian countries since late 2003 has led to the application of various techniques to diagnose the disease. For the isolation technique, influenza virus progenies released from the infected cells are primarily recognized by hemagglutination (HA) test. In addition, HA is employed to quantify the amount of hemagglutinin antigen used in HA inhibition (HI) assay (7). Importantly, the World Health Organization (WHO) has specified the presence of a horse erythrocyte HI titer of ≥160 in adjunct with a microneutralization (microNT) antibody titer of ≥80 in a single serum collected at day 14 or later as one among other criteria for a confirmed case of H5N1 infection (8).
Agglutination of erythrocytes by influenza viruses is mediated by the interaction between the receptor binding site (RBS) in hemagglutinin molecule and the sialyl receptor. Human influenza H1N1 and H3N2 viruses preferentially bind to a sialic acid receptor, the oligosaccharide side chain of which is linked with α2,6-galactose linkage (SAα2,6Gal), while avian and equine influenza viruses prefer an α2,3-galactose linkage (SAα2,3Gal). Horse and cow erythrocytes contain mainly an SAα2,3Gal linkage but no SAα2,6Gal (1). Chicken and goose erythrocytes contain more SAα2,3Gal linkage than SAα2,6Gal, while this is reversed with human O cells and pig, guinea pig, and turkey erythrocytes (1, 2).
Herein, five erythrocyte species (horse, goose, chicken, guinea pig and human O cells) were tested by HA assay against 14 H5N1 clade 1 isolates from Thailand, including five from humans, seven from wild and domestic birds, one from a tiger and, one from a clouded leopard, together with one human H1N1 isolate and five H3N2 isolates. Final concentrations of 0.5% goose, 0.5% chicken, 0.75% guinea pig, 0.75% human group O, and 1% horse erythrocytes were used. Except for those from the horse, erythrocytes were suspended in phosphate-buffered saline, pH 7.2. Horse erythrocytes were suspended in phosphate-buffered saline plus 0.5% bovine serum albumin. These protocols were followed as described previously (6, 7). A reaction well, consisting of 50 μl of diluted virus and 50 μl of erythrocyte suspensions, was incubated for 1 h at 4°C before the agglutination pattern was read. One HA unit is defined as the highest virus dilution that yields complete HA.
The experiments demonstrated that 13 of 14 H5N1 isolates could agglutinate erythrocytes from all five species with a statistical difference in the extent of titer (Friedman test, P < 0.05). Interestingly, an isolate from the clouded leopard could not agglutinate horse erythrocytes (Table 1). The result was consistent, as repeatedly tested with erythrocytes from three donors within one species. Goose erythrocytes yielded the highest HA titer, followed in order of sensitivity by chicken, guinea pig, human, and horse erythrocytes (Wilcoxon's signed-rank test, P < 0.005).
TABLE 1.
HA titers of influenza A viruses as assayed with erythrocytes from different species
| Influenza A virus and HA GMT | Passage history | Accession no. | HA titer by erythrocyte sourcea
|
||||
|---|---|---|---|---|---|---|---|
| Goose | Chicken | Guinea pig | Human | Horse | |||
| H1N1 human influenza virus | |||||||
| A/New Caledonia/20/99-like virus (Siriraj 07/00) | MDCK7 | EF568930 | 128 | 64 | 128 | 64 | <2 |
| H3N2 human influenza viruses | |||||||
| A/Sydney/05/97-like virus (Siriraj 08/98) | MDCK4 | EF568929 | 64 | 64 | 128 | 64 | <2 |
| A/Fujian/411/02-like virus (Siriraj 03/04) | MDCK6 | EF568924 | 32 | <2 | 64 | 64 | <2 |
| A/Fujian/411/02-like virus (Siriraj 01/03) | MDCK3 | EF568925 | 32 | <2 | 32 | 32 | <2 |
| A/Fujian/411/02-like virus (Siriraj 02/03) | MDCK2 | EF568926 | 4 | <2 | 32 | 32 | <2 |
| A/California/07/04-like virus (Siriraj 12/04) | MDCK4 | EF568927 | 4 | <2 | 16 | 32 | <2 |
| H5N1 influenza viruses | |||||||
| A/Thailand/1(KAN-1)/04 | LLC-MK2, MDCK8 | AY555150 | 512 | 512 | 128 | 128 | 128 |
| A/Thailand/2(SP-33)/04 | MDCK6 | AY555153 | 64 | 64 | 64 | 32 | 16 |
| A/Thailand/3(SP-83)/04 | MDCK5 | AY577314 | 64 | 64 | 16 | 32 | 16 |
| A/Thailand/5(KK-494)/04 | MDCK4 | AY627885 | 128 | 64 | 32 | 64 | 32 |
| A/Thailand/676(NYK)/05 | MDCK9 | DQ360835 | 256 | 256 | 512 | 256 | 32 |
| A/Great Barbet/Thailand/ VSMU-2-CBI/2005 | MDCK2 | EF206697 | 128 | 128 | 64 | 64 | 32 |
| A/Green Peafowl/Thailand/ VSMU-3-CBI/2005 | MDCK2 | EF206700 | 256 | 128 | 64 | 64 | 64 |
| A/Gray-Crowed Crane/ Thailand/VSMU-4-CBI/2005 | MDCK2 | EF206696 | 512 | 256 | 128 | 128 | 128 |
| A/Tree Sparrow/Thailand/ VSMU-16-RBR/2005 | MDCK4 | EF178506 | 64 | 64 | 32 | 32 | 16 |
| A/Golden Pheasant/Thailand/ VSMU-21-SPB/2005 | MDCK1 | EF178517 | 128 | 64 | 32 | 16 | 32 |
| A/Pigeon/Thailand/VSMU-25-BKK/2005 | MDCK3 | EF206698 | 64 | 64 | 32 | 16 | 16 |
| A/Chicken/Thailand (Suphanburi)/137/05 | MDCK4 | EF568922 | 512 | 512 | 64 | 64 | 128 |
| A/Tiger/Thailand/VSMU-11-SPB/2004 | Egg 1 | EF178531 | 512 | 512 | 256 | 256 | 256 |
| A/Clouded Leopard/Thailand (Chonburi)/AI-1216A/2004 | Egg 1 | EF568923 | 512 | 512 | 256 | 256 | <2 |
| HA GMT of H5N1 viruses | 190.21 | 156.03 | 74.25 | 67.25 | 33.62 | ||
The most frequent titer obtained from three erythrocyte donors within the same species is presented.
Hemagglutinin amino acid sequences of our H5N1 isolates were compared with those of A/Goose/Guangdong/1/96 (the ancestor) and with Hong Kong virus 1997 (5) (see Fig. S1 in the supplemental material). No change in RBS was found, except for one isolate, A/Thailand/676(NYK)/05, which contained a mutational change A134V in RBS. However, this mutational change did not relate to erythrocyte binding preference. It remains to be elucidated why the isolate from the clouded leopard could not agglutinate horse erythrocytes while there were no change in RBS and no difference in the deduced amino sequence of hemagglutinin. Receptor specificity of influenza viruses is influenced by both the galactose linkage and species of sialic acid: N-acetylneuraminic acid (NeuAc) or N-glycolylneuramic acid (NeuGc). Horse erythrocytes contained only NeuGcα2,3Gal (1, 3). Therefore, loss of the ability to agglutinate horse erythrocytes may be related to loss of the ability to recognize either NeuGc or galactose linkage (1, 3).
The study of human influenza H1N1 and H3N2 viruses showed that all six isolates could not agglutinate horse erythrocytes (Table 1). An A/Sydney/05/97(H3N2)-like isolate could agglutinate chicken erythrocytes, while all three A/Fujian/411/02(H3N2)-like isolates and one A/California/07/04(H3N2)-like isolate could not. This result supported previous findings that current H3N2 isolates have lost their agglutinating activity with chicken erythrocytes (2, 4). However, all of the data, including ours, were discrete and could not conclude that there was a correlation between certain mutational positions and loss of the agglutinating activity (data not shown).
Fourteen serum samples from seven H5N1 patients, including the survivors, were assayed for H5N1 antibody. Two H5N1 strains, A/Thailand/1(KAN-1)/04 which was the first human isolate from the country (5), and A/Thailand/676(NYK)/05, as described above, were selected as the test antigens. This study was approved by the Institution Ethical Committee for Human Research.
In the HI test, serum was pretreated with a receptor-destroying enzyme (Denka Seiken, Japan) at final dilution of 1:4 for 16 h at 37°C followed by heat inactivation for 30 min at 56°C and absorbed with a 50% erythrocyte suspension for 60 min at 4°C. A mixture of 25 μl of the diluted serum and 25 μl of the test virus at a concentration of 4 HA units was incubated for 30 min at room temperature before addition of 50 μl of erythrocyte suspension. The end result was read after incubation for 1 h at 4°C. HI antibody titer is defined as the final serum dilution that completely inhibits HA. Again, three donors from each of the five erythrocyte species were tested in separate runs with consistent results.
The results demonstrated that horse erythrocytes, which were the least sensitive in HA, gave the highest geometric mean titer (GMT) of antibody when A/Thailand/1(KAN-1)/04 was used as the test antigen (Wilcoxon's signed-rank test, P < 0.005). Goose erythrocytes were ranked second, followed by human, guinea pig, and chicken erythrocytes. In contrast, when A/Thailand/676(NYK)/05 was used as the test antigen, goose erythrocytes yielded the highest GMT, followed in order by chicken, horse, human, and guinea pig erythrocytes. However, a statistically significant difference was not found (Wilcoxon's signed-rank test, P > 0.005) (Table 2). Collectively, the level of HI antibody titer was dependent on both the erythrocyte species and the test antigen used. We also showed that HI is more sensitive for strain differentiation than microNT.
TABLE 2.
H5N1 HI antibody titers as tested by different erythrocyte species
| Patient, sequential sample no., and HI antibody GMT | Time after onset of illness | A/Thailand/1(KAN-1)/04(H5N1)a
|
A/Thailand/676(NYK)/05(H5N1)b
|
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HI antibody titer by erythrocyte source
|
NT antibody titer | HI antibody titer by erythrocyte source
|
NT antibody titer | ||||||||||
| Goose | Chicken | Guinea pig | Human | Horse | Goose | Chicken | Guinea pig | Human | Horse | ||||
| Patient 1 | |||||||||||||
| 1 | 8 days | 20 | <20 | 20 | <20 | 20 | <5 | 20 | <20 | <20 | <20 | 20 | |
| 2 | 17 days | 640 | 640 | 640 | 640 | 1,280 | 1,280 | 1,280 | 1,280 | 1,280 | 1,280 | 640 | |
| Patient 2 | 2 yr, 2 mo | 20 | <20 | 20 | 20 | 40 | 80 | 160 | 80 | 80 | 80 | 80 | 40 |
| Patient 3 | 2 yr, 1 mo | 20 | <20 | 20 | 20 | 40 | 160 | 160 | 80 | 80 | 80 | 80 | 80 |
| Patient 4 | |||||||||||||
| 1 | 10 days | 20 | <20 | <20 | 20 | 20 | <5 | ||||||
| 2 | 12 days | 20 | <20 | 20 | <20 | 20 | 5 | 20 | 20 | 20 | |||
| 3 | 6 mo | 40 | 20 | 20 | 40 | 80 | 160 | 160 | 320 | 160 | 80 | 320 | |
| 4 | 1 year | 40 | 20 | 20 | 40 | 80 | 160 | 160 | 160 | 80 | 160 | 160 | 80 |
| Patient 5 | 7 days | <20 | <20 | <20 | <20 | <20 | <5 | 20 | <20 | <20 | <20 | 20 | |
| Patient 6 | |||||||||||||
| 1 | 4 days | <20 | <20 | <20 | <20 | <20 | 5 | <20 | <20 | <20 | <20 | <20 | |
| 2 | 15 days | 320 | 160 | 160 | 160 | 640 | 1,280 | 1,280 | 1,280 | 1,280 | 1,280 | 1,280 | |
| 3 | 21 days | 160 | 80 | 80 | 160 | 320 | 640 | 640 | 1280 | 640 | 640 | 640 | |
| 4 | 5 mo, 9 days | 20 | 20 | 20 | 20 | 40 | 80 | 160 | 80 | 80 | 40 | 40 | 80 |
| Patient 7 | 5 days | 80 | 40 | 80 | 160 | 320 | 320 | 640 | 640 | 320 | 320 | 640 | |
| GMT of HI antibody | 40.00 | 24.38 | 31.23 | 36.26 | 65.63 | 136.35 | 134.54 | 93.88 | 100.79 | 110.16 | |||
A/Thailand/1(KAN-1)/04 was isolated from patient 1.
A/Thailand/676(NYK)/05 was isolated from patient 5.
Our study proposes that goose erythrocytes confer a greater advantage for recognition of H5N1 viruses and HI antibody assay. Whether this finding is also generalized for H5N1 clade 2 viruses need to be investigated.
Supplementary Material
Acknowledgments
We would like to thank the National Center for Genetic Engineering and Biotechnology (BIOTEC), Thailand, and the Thailand Research Fund for Advanced Research Scholar for support.
We also thank Stephen Durako, Westat, for manuscript review and Jarunee Prasertsophon, Kannika Nateerom, and Rasameepen Po-ngakern for laboratory support.
Footnotes
Published ahead of print on 23 May 2007.
Supplemental material for this article may be found at http://jcm.asm.org/.
REFERENCES
- 1.Ito, T., Y. Suzuki, L. Mitnaul, A. Vines, H. Kida, and Y. Kawaoka. 1997. Receptor specificity of influenza A viruses correlates with the agglutination of erythrocytes from different animal species. Virology 227:493-499. [DOI] [PubMed] [Google Scholar]
- 2.Medeiros, R., N. Escriou, N. Naffakh, J.-C. Manuguerra, and S. van der Werf. 2001. Hemagglutinin residues of recent human A (H3N2) influenza viruses that contribute to the inability to agglutinate chicken erythrocytes. Virology 289:74-85. [DOI] [PubMed] [Google Scholar]
- 3.Neumann, G., and Y. Kawaoka. 2006. Host range restriction and pathogenicity in the context of influenza pandemic. Emerg. Infect. Dis. 12:881-886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nobusawa, E., H. Ishihara, T. Morishita, K. Sato, and K. Nakajima. 2000. Change in receptor-binding specificity of recent human influenza A viruses (H3N2): a single amino acid change in hemagglutinin altered its recognition of sialyloligosaccharides. Virology 278:587-596. [DOI] [PubMed] [Google Scholar]
- 5.Puthavathana, P., P. Auewarakul, P. C. Charoenying, K. Sangsiriwut, P. Pooruk, K. Boonnak, R. Khanyok, P. Thawachsupa, R. Kijphati, and P. Sawanpanyalert. 2005. Molecular characterization of the complete genome of human influenza H5N1 virus isolates from Thailand. J. Gen. Virol. 86:423-433. [DOI] [PubMed] [Google Scholar]
- 6.Stephenson, I., J. M. Wood, K. G. Nicholson, and M. C. Zambon. 2003. Sialic acid receptor specificity on erythrocytes affects detection of antibody to avian influenza haemagglutinin. J. Med. Virol. 70:391-398. [DOI] [PubMed] [Google Scholar]
- 7.World Health Organization. 2002. WHO manual on animal influenza diagnosis and surveillance. WHO/CDS/CSR/NCS/2002.5. http://www.who.int/csr/resources/publications/influenza/en/whocdscsrncs20025rev.pdf. Accessed 24 November 2006.
- 8.World Health Organization. 2006. WHO case definitions for human infections with influenza A (H5N1) virus, 29 August 2006. http://www.who.int/csr/disease/avian_influenza/guidelines/case_definition2006_08_29/en/print.html. Accessed 24 November 2006.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.