Some H5N1 avian influenza viruses (Claas et al., 1998
; Li et al., 2004
) exhibit high pathogenicity in humans with pandemic potential (Gambotto et al., 2008
). This underscores the need to better understand the mechanisms that determine the ability of emerging influenza viruses to cross the species barrier into humans. This process is regulated in part by the viral haemagglutinin (HA) and host sialic acid receptors (Suzuki et al., 2000
). The viral polymerase complex is also recognized to play an important role in IAV host-cell specificity and pathogenicity (Gabriel et al., 2005
; Labadie et al., 2007
; Massin et al., 2001
; Salomon et al., 2006
), but the underlying basis for this remains poorly understood.
In this report, we show that the IAV 3P complex can be successfully purified from human lung epithelial cells using an adenoviral system. We analysed the functional activity of IAV polymerase complexes from the WSN (H1N1) and VN1203 (H5N1) strains, as well as chimeric complexes in which the PB2 subunits were exchanged. Our data show that the H5N1 polymerase complex has more robust thermotolerance than its WSN counterpart, which is active over a narrow temperature range (30–34 °C) with a more than 50
% reduced activity at 37 °C. This effect can be attributed to the PB2 subunit, since a simple exchange of this protein was associated with substantial changes in the temperature sensitivity of the polymerase activity. This effect does not depend on the presence of a glutamic acid residue at position 627, since the PB2 subunits from both the VN1203 and the WSN strain harbour a lysine at position 627 (K627
). This suggests that additional determinants within PB2 contribute to virus host-range and replication efficiency as well (Gabriel et al., 2005
; Naffakh et al., 2000
; Shinya et al., 2004
We also generated complexes in which we substituted the PB2 subunit in the WSN 3P complex with an avian (H3N2) PB2 counterpart. Functional analysis of these complexes revealed that, like the H5N1 PB2 subunit, the avian PB2 subunit also conferred improved thermotolerance on the WSN 3P complex. Associated host-factors could possibly contribute to the functional properties of the purified 3P complex, although future experiments will be needed to address this question.
Our analyses also showed that the E627
residue in the avian PB2 was not required for the enhanced functional activity of the chimeric W/W/N. The E627K mutation in the avian PB2 led to only a slight increase in thermotolerance. Studies that have described the role of PB2 amino acid residue 627 as a determinant of cold sensitivity of avian influenza viruses were performed using transcription–replication experiments (Labadie et al., 2007
; Massin et al., 2001
) or by studying the replication of infectious, recombinant viruses (Hatta et al., 2007
; Salomon et al., 2006
). These assays effectively examine both RNA transcription and genome replication, since a defect in either process would lead to a reduction in measured overall replication. In contrast, our experiments examined only viral transcription. Since viral transcription and replication occur via distinct mechanisms, our findings suggest that PB2 residues other than amino acid 627 may regulate temperature sensitivity of viral transcription – and that mutations at these residues may contribute to virus-host adaptation and temperature sensitivity of overall virus replication, through effects on RNA transcription (Chen et al., 2006
; Finkelstein et al., 2007
). Further studies to address this hypothesis are ongoing.
At the mechanistic level, our results show that the thermotolerance-enhancing activity of the avian PB2 can be attributed to an increase in functional stability of the polymerase complex. Brownlee & Sharps (2002)
have speculated that, in the absence of the vRNA promoter, the WSN polymerase complex is open and labile to heat inactivation. We showed that the half-life of the WSN complex in the absence of vRNA was roughly 10 min at 30 °C, whereas the functional half-life of the W/W/N complex exceeded 2 h. The short functional half-life of the WSN complex was not due to disintegration or disassembly of the tripartite polymerase protein complex, since the complex remained intact for at least 1 h.
Our findings support the earlier hypothesis (Brownlee & Sharps, 2002
) that avian IAV RNA polymerases may have increased heat stability, due to the need of avian viruses to replicate at elevated temperatures present in birds. Brownlee also hypothesized that highly pathogenic human influenza viruses may possess polymerases with increased thermostability, relative to less pathogenic strains. Our results provide support for both of these predictions, and suggest the need to examine the thermostability of RNA polymerases from additional viruses.
In conclusion, the results reported in this paper define a new and previously unappreciated role for the influenza PB2 protein in determining the thermostability of the virus RNA polymerase. Future studies will determine the specific amino acid residues that contribute to this property, and will assess whether increased thermostability is associated with greater pathogenicity and/or host adaptation by emerging influenza viruses.