Main

The presence of H5N1 avian influenza virus in Africa was first reported by the United Nations Food and Agriculture Organization in commercial poultry farms in Kaduna state in northern Nigeria on 7 February 2006. The poultry farming industry is second only to oil production in Nigeria and is particularly vulnerable to the introduction of infectious agents because chickens are imported from all over the world without rigorous biosecurity safeguards. This vulnerability is increased because the country has many bird sanctuaries along the two flight paths that link Nigeria with the southern Russian region and Europe, and with western Asia1. However, a sero-epidemiological study2 of 65 flocks in the southwestern states that form the hub of the Nigerian poultry industry showed no evidence of avian influenza in poultry up to the time the study was terminated in May 2004.

We have now recovered virus isolates that tested positive in a generic test3 for influenza A from 18 cloacal swabs taken from poultry from two farms (‘SO’ on 24 February 2006 and ‘BA’ on 8 March 2006) in Lagos state in the far southwest of the country (for map and methods, see supplementary information). Sequencing of these isolates confirmed the presence of H5N1. The viral haemagglutinin cleavage-site sequence PQGERRRKKRG (single-letter amino-acid notation) was identical to that of the highly pathogenic strains already found in Europe, Russia and central Asia. Figure 1 shows that representative sequences from the complete haemagglutinin gene cluster with those of the 2005–06 H5N1 strains from western Asia (Mongolia; Qinghai in China; Crimea in the Ukraine; Kurgan, Astrakhan and Novosibirsk in Russia; Iran; Iraq), Europe (Turkey; Croatia; France; Italy; Germany) and Africa (Egypt).

Figure 1: H5N1 in Nigerian poultry.
figure 1

Phylogeny of the viral haemagglutinin gene in the Nigerian lineages and in the most closely related strains. The maximum likelihood method was used, with estimated transition/transversion ratio and a gamma-distribution substitution model and 100 bootstrap values. Bootstrap values of the principal nodes are shown. A/chicken/Hebei/2005 was used as an outgroup to build the tree. Countries having their first outbreak in 2006 are shown in blue; west Asian and Russian strains are in yellow and orange, respectively; southeast Asian strains are in green; and Nigerian strains are in red.

Full size image

The percentage variation in nucleotide sequence between the strains from the two Lagos state farms in the southwest and DQ406728, the strain from northern Nigeria, was 0.79%, 0.87% and 1.14%, respectively. The complete haemagglutinin sequences from the BA, SO and northern Nigerian isolates resembled most closely the A/Cygnus olor/Astrakhan/Ast05-2/2005, A/chicken/Egypt/960N3-004/2006 and A/chicken/Kurgan/3/2005 strains, respectively (Fig. 1).

The BA sequences were even more closely related to sequences from swans from Poland and Germany and a buzzard from Denmark (0.3% diversity on the haemagglutinin gene; Ian Brown, personal communication) and, despite the relative proximity of the SO and BA farms (less than 50 kilometres apart), strains from the BA farm were more closely related to the northern Nigerian strain than to the SO farm strains (see supplementary information). We conclude that the highly pathogenic avian influenza virus strains in northern Nigeria and in Lagos state belong to three distinct genetic lineages and were therefore probably introduced independently. Our sequence data rule out a southeast-Asian origin for these different lineages.

By 20 February 2006, the Nigerian federal government had set up a biosafety zone to contain the virus in the north, whose efficiency was questioned when the virus was detected in Lagos state. Our results indicate, however, that the viruses in the southwest were not introduced from the north, but that instead they all arrived independently, perhaps at inland waters and key bird areas or through unprotected trade. Meanwhile, the H5N1 virus has reached 14 of the 31 federal states of Nigeria and it is difficult to see how the virus might still be contained without depopulation measures, combined with surveillance and large-scale vaccination.