Highly pathogenic avian influenza viruses (HPAIV) cause devastating outbreaks in domestic poultry worldwide. Moreover, they proved to be responsible for repeated severe human infections with high lethality raising concerns about their pandemic potential. HPAIV are restricted to the hemagglutinin serotypes H5 and H7 and evolve from low-pathogenic precursors. Since the introduction of an artificial polybasic HA cleavage site into non H5/H7 HA of low-pathogenic strains may result in high-pathogenic viruses, the emergence of HPAIV with novel serotypes is conceivable. However, the observed HA serotype restriction in “natural” HPAIV indicates a unique predisposition of the H5/H7 HA for an insertion mutation coding for a polybasic hemagglutinin cleavage site.
Highly pathogenic avian influenza viruses (HPAIV), formerly known as fowl plague viruses, cause devastating losses in domestic poultry. Within the past 130 years, more than 30 outbreaks have been reported worldwide. Furthermore, repeated severe HPAIV infections in humans with high lethality have raised concerns about adaptation to the human host sparking a novel pandemic. An essential step in the replication cycle of all influenza A viruses is the cleavage of the hemagglutinin (HA) precursor HA0 into the HA1 and HA2 fragments. Only the cleaved HA can undergo a conformational change in the acidic endosome resulting in exposure of the hydrophobic N termini of the HA2 fragments (the fusion peptides) to mediate fusion of the virion envelope with the endosomal membrane. This proteolytic activation requires host proteases that recognize specific cleavage sites. The common HA cleavage site (HACS) contains a monobasic motif that is prevalent in all human, mammalian and low-pathogenic avian influenza viruses (LPAIV). In contrast, all HPAIV carry a polybasic HACS. This difference in cleavage motifs leads to the recruitment of different host proteases. The monobasic HACS is cleaved by tissue-restricted proteases such as TMPRSS2, HAT, other trypsin-like and perhaps yet unidentified proteases, whereas the polybasic HACS is processed by the ubiquitous proprotein convertase furin and related subtilases. This utilization of different proteases results in distinct tissue tropisms. Whereas LPAIV and human influenza viruses are mostly confined to the digestive or respiratory tracts, respectively, the HPAIV display a very broad organ tropism affecting several organs including the central nervous system, often considered the prime cause of death in infected birds. Since experimental removal of the polybasic HACS invariably resulted in a virus with low-pathogenic phenotype, it is considered the key virulence determinant of HPAIV.
All HPAIV have evolved from low-pathogenic precursors due to extension of the HACS by an insertion mutation of the HA gene. Remarkably, different amino acid sequences at the HACS indicate independent mutation events. For a few HPAIV, the insert was found to originate from recombination with viral (NP or M gene) or host (ribosomal 28S) RNA, or was suggested to result from polymerase slippage; however, in most strains, the mechanism for insertion mutation has remained unknown.
Beside the polybasic HACS, HPAIV differ from LPAIV in that they are restricted to the HA serotypes H5 and H7, two of 16 serotypes known in avian viruses. Since the reason for this serotype restriction is unclear, we aimed to distinguish between the compatibility of a polybasic cleavage site with H5/H7 HA only and a unique predisposition of those two serotypes for insertion mutations. Artificial introduction of a polybasic HACS into low-pathogenic strains of HA serotypes H3, H4 or H9 did not lead to a high-pathogenic phenotype, in contrast to an HACS mutant of an H6N1 virus. Remarkably, a polybasic cleavage site mutant of a low-pathogenic H5N1 virus did not display high virulence either. Reassortants carrying the HA gene from an H5 HPAIV with the other seven genes originating from either the H9 or H5 LPAIV showed only temporary non-lethal disease in chicken, whereas reassortants with the reciprocal gene constellation, i.e., the LPAIV H9 or H5 HA with artificial polybasic HACS and the genetic background from an HPAIV, were lethal for chickens. Therefore, virulence determinants of HPAIV exist beyond the essential polybasic HACS and reside both in the HA and the other gene segments.
To investigate whether HPAIV of other serotypes could emerge in the genetic background of an HPAIV, we introduced a polybasic cleavage site into the HA of LPAIV with serotypes H1, H2, H3, H4, H6, H8, H10, H11, H14 or H15, and rescued HA reassortants after co-transfection with the genes from either a low-pathogenic H9N2 or high-pathogenic H5N1 avian strain. Those recombinants which contained the engineered H2, H4, H8 or H14 genes in the HPAIV background, were lethal in chickens and exhibited intravenous pathogenicity indices of 2.79, 2.37, 2.85 and 2.61, respectively, equivalent to naturally occurring H5 or H7 HPAIV. Thus, in the presence of a polybasic HACS, nonH5/H7 HA can support a highly pathogenic phenotype in the appropriate viral background, indicating requirement for further adaptation. Hence, the observed restriction of natural HPAIV to serotypes H5 and H7 is likely due to their unique predisposition for acquisition of a polybasic HACS.
In the light of our finding that HPAIV with non-H5/H7 serotypes can be generated experimentally, it is conceivable that HPAIV with novel serotypes could emerge in the future. H5N1 HPAIV have become endemic in several regions such as Southeast Asia or Egypt yielding ample opportunity for reassortment with an LPAIV to evolve into an HPAIV. However, the molecular basis for the emergence of an HPAIV is still not clear and may include these three preconditions: the HA has to be prone to acquire a polybasic HACS, a lowered optimal pH to trigger fusion of the HA has to be achieved, and other virulence and host range determinants like the NA stalk deletion have to be established. Furthermore, it is still uncertain whether such additional virulence determinants have to be accumulated before or after formation of the polybasic HACS. Moreover, in which avian host does this process take place? This evolutionary step is likely to happen in gallinaceous poultry prior to an HPAI outbreak as suggested by previous observations. For instance, prolonged circulation of low-pathogenic precursor viruses in chickens was found for several months before the outbreaks in Mexico 1994 and Italy 1999. Therefore, at least some of the virulence or host range determinants might be required for virus survival in the chicken flocks prior to acquisition of the polybasic HACS by an insertion mutation that occurs at considerable lower probabilities than point mutations.
Taken together, the natural formation of a polybasic HACS has so far been restricted to H5 or H7 LPAIV. This HA serotype restriction observed in authentic HPAIV points to a unique predisposition of the H5/H7 HA for polybasic HACS formation. Since in presence of the required virulence determinants in the HA and other genes, the experimental introduction of a polybasic HACS may result in high-pathogenic viruses with HA serotypes other than H5 or H7, the emergence of HPAIV with novel serotypes is conceivable.