Since 2009, Egypt has reported a higher number of HPAI (H5N1) virus infections in humans than any other country. At the end of 2011, Egypt had reported 39 (63%) of the total 62 human cases in the world for that year, placing Egypt second only to Indonesia in the number of reported human infections since 2003 (7
). Exposure to sick or dead poultry has been reported as the likely source of infection for nearly all human cases in Egypt (29
). Most of those exposures were described as occurring in backyard poultry settings (although infection in industrial/commercial settings could not be ruled out if someone worked in or visited these settings) (29
). The contrast between the rising numbers of human infections detected each year in Egypt and the declining case-fatality ratios since 2009 led to speculation about the evolution of new virologic properties influencing the epidemiology of subtype H5N1 virus in Egypt (2
). To investigate possible molecular epidemiologic correlates of these trends, we analyzed the complete genomes of viruses isolated from humans in Egypt during 2007–2011.
Phylogenetic analysis of HA genes indicated that subtype H5N1 viruses from clades 2.2.1 and 184.108.40.206 continued to co-circulate in recent years (6
). In addition, the HA of clade 2.2.1 viruses was also found to cluster in 1 of 4 distinct monophyletic groups (previously termed groups A–D) (8
). Of the human infections during 2009–2011, 95% were caused by viruses from a single phylogenetic group, clade 2.2.1-C, whereas infections detected during 2007–2008 involved viruses from 2.2.1-B and -D.
In contrast with the multiple human infections caused by the clade 2.2.1 viruses, only 1 human infection by a clade 220.127.116.11 virus was detected (). The predominance of clade 2.2.1-C viruses among zoonotic infections in humans appears to be associated with the persistent circulation of this group of viruses in backyard or peridomestic poultry (30). However, there remains some evolutionary divergence between 2 discrete HA clusters detected in 2011 that may correlate with the geographic separation between the majority of group C genes that originated from the Nile Delta and those from strain A/Egypt/N0423/2011, which was collected from the Aswan governorate in the south of the country. The paucity of human infections with clade 18.104.22.168 viruses, which circulate predominantly in commercial poultry (30
), may result from intrinsic viral properties or from husbandry practices that reduce the probability of zoonotic infection. Thus, exposure to subtype H5N1 virus in backyard or peridomestic environments, rather than commercial settings, appears to be correlated with greater risk for zoonotic H5N1 infections.
The genetic variation that was observed among the HA genes of clade 2.2.1 viruses correlated with an increased complexity in the antigenic characteristics of the viruses currently circulating in Egypt. Although all clade 2.2.1 viruses remain crossreactive to sera produced against reference viruses from the same clade, HI assay results indicated that continued variation among A, B, C, and D viruses has resulted in reduced titers of recent 2.2.1-C viruses against antisera to older clade 2.2.1 viruses. It should be noted, that the recent 2011 isolates were antigenically closely related to the proposed WHO candidate vaccine virus, A/Egypt/N03072/2010, indicating a good antigenic match between currently circulating strains and the proposed vaccine.
Although the HA and NA genes play a major role in the transmission of HPAI (H5N1) virus, the internal genes can also modulate pathogenicity and transmissibility of the virus (17
). To further investigate the evolution of the internal genes of subtype H5N1 viruses from humans and detect possible reassortment (intraclade or other), we performed a systematic analysis of the phylogenetic relationships among all the genes. There was a notable difference in the topology of the phylogenetic tree of the NS gene compared with other internal genes (supported by bootstrap values >80) in that no distinct 2.2.1-B and 2.2.1-D subgroups were evident for the NS gene tree. All the internal genes that did not appear to co-evolve with their surface genes show the closest relationship to the genes of A/turkey/Turkey/2005, indicative of a lack of strong selective pressure. These findings also confirmed previous reports that subtype H5N1 genes have not been introduced from Asia into Egypt since 2006. The phylogenies of all the genes from the viruses in Egypt lacked evidence of reassortment with other avian influenza genes.
The co-circulation of H9N2 and potentially other LPAI A viruses in Egypt has been reported (31
); thus, the absence of genetic reassortment between subtype H5N1 and LPAI avian viruses from poultry could be considered unexpected. In contrast to reassortant genotypes that have been identified in other countries, the homogeneous genetic makeup of subtype H5N1 viruses in Egypt may stem from the characteristics of the poultry trade with neighboring countries or from other factors leading to fewer opportunities for co-infections with other avian influenza viruses.
At the protein level, evolution among 2.2.1-C viruses, compared with that among other clade 2.2.1 viruses, has resulted in the fixation of at least 4 substitutions in the HA protein. When found together, 2 of the 4 HA mutations have been implicated as possible host adaptation markers: the deletion of residue 129 and the I151T substitution enabled in vitro binding to α-2,6 sialosides and virulence in mice (25
). However, these 2 mutations may also mediate antigenic drift (20
). The K615R substitution in the PA proteins of 2.2.1-C viruses represents another potential marker of host adaptation (17
). No markers of antiviral drug resistance were found to be conserved in either the NA or M2 of any of the human-derived subtype H5N1 viruses. Epidemiologic investigations of the human cases of subtype H5N1 infection from which the study isolates were derived could not confirm which patients had received oseltamivir or other antiviral treatments. However, previous analyses of subtype H5N1 viruses isolated from poultry in Egypt were found to have M2 and NA proteins with antiviral resistance markers, indicating that these mutations exist to some extent in viruses circulating within the poultry population (8
In summary, our findings indicate that the recent subtype H5N1 viruses isolated from human infections originated from poultry. These viruses evolved from a single genotype introduced into Egypt in 2005/2006; there is no evidence of subsequent reassortment with new subtype H5N1 virus genes introduced into Egypt or resident LPAI viruses. The viruses have been observed in all subtype H5N1 infections in humans since 2009 and belong to a subgroup, termed 2.2.1-C, with unique genetic signatures that may contribute to their persistence in poultry. A rationale for linking the observed amino acid changes to the decline in case-fatality ratios since 2009 could not be identified. Systematic analysis of subtype H5N1 viruses in Egypt is critical to better understand the genetic and phenotypic evolution of subtype H5N1 viruses in Egypt and to inform public health programs to reduce the risk for zoonotic infections and prevent or mitigate a potential pandemic.