Monolayer cultures of 293T and MDCK cells were maintained at 37°C with 5% CO₂ in Dulbecco's Modified Eagle Medium (DMEM) and minimal essential medium (MEM) (Nissui), respectively, supplemented with 10% fetal bovine serum (Bovogen). Influenza virus strain A/WSN/33 (WSN) was prepared as previously described (Kawaguchi et al., 2005). Cycloheximide (CHX) was purchased from Sigma-Aldrich.

To construct plasmids from which human DNA-dependent RNA polymerase I (Pol I) transcribes mutated vRNAs, we amplified fragments containing mutated segment 2 by PCR using a plasmid containing wild type WSN segments in pHH21 vector (Neumann et al., 1999) as a template with sets of phosphorylated primers (see Table S1 in the supplemental material). The amplified PCR products were self-ligated followed by sequencing. To generate recombinant viruses containing viral RNAs of WSN and mutated segment 2, reverse genetics system was used as described previously (Neumann et al., 1999). After 48 h post transfection (hpt), aliquots of cell culture supernatants were used for virus amplification in MDCK cells. At 48 h post infection (hpi), the culture fluid was collected, and the virus titer of these recombinant viruses was determined by plaque assays.

MDCK cells were infected with recombinant viruses at the multiplicity of infection (MOI) of 2.5. At 9 hpi, total RNA was isolated by the acid guanidine-phenol-chloroform method. To measure the accumulation levels of viral mRNA, cRNA, and vRNA, quantitative RT-PCR (qRT-PCR) was performed. Total RNAs were subjected to reverse transcription using ReverTraAce (Toyobo) with either (i) oligo (dT)₂₀, (ii) 5′-AGTAGAAACAAGGGTATTTTTCTTTA-3′, or (iii) 5′-GACGATGCAACGGCTGGTCTG-3′ for synthesizing cDNA from segment 5 mRNA, cRNA, and vRNA, respectively (Kawaguchi and Nagata, 2007; Sugiyama et al., 2009). The synthesized single-stranded cDNAs were subjected to real-time quantitative PCR analysis (Thermal Cycler Dice real-time system TP800; TaKaRa) with SYBR Premix Ex Taq (TaKaRa) and a set of specific primers for segment 5 cDNA (see supplementary methods). The levels of these RNAs were normalized by the amount of cellular β-actin mRNA measured using specific primers (see supplementary methods). These results are averages from three independent experiments with standard deviations. The level of significance was determined by Student's t-test (unpaired).

The N-terminal region of PB1 (1–83 a.a.) contains the PA binding site and both putative vRNA and cRNA promoter binding sites (Figure 1A). It is shown by alignment of amino acid sequences that the PB1 region between a.a. positions 1–50 are highly conserved between influenza A and B viruses, while the region between a.a. positions 51–83 differ between two viruses (Figure 1B). In the highly conserved region, except for the PA binding site, a.a. at positions 16, 27, and 44 are different between these viruses. Furthermore, these a.a. positions in PB1 are conserved more than 99% of influenza A and B viruses deposited in the NCBI Influenza Virus Sequence Database (Table 1). To elucidate the functional importance of these a.a. for viral RNA synthesis, we generated influenza A viruses containing Ala at the a.a. position 16 (N16A), Val at the a.a. position 27 (D27V), and Ile at the a.a. position 44 (N44I) by reverse genetics. We examined the RdRp activity by measuring the accumulation levels of viral mRNA, cRNA, and vRNA by qRT-PCR (Figure 2). The levels of all three type RNAs from D27V were increased compared with those from wild type and N16A virus, while those from N44I were significantly decreased. Based on the result that mutations at the positions 27 and 44 affect the synthesis activity of all viral RNAs equally, there could be two possibilities: (i) these mutations affect on the vRNA promoter recognition and followed by cRNA/mRNA synthesis, but do not affect on the cRNA promoter recognition and followed by vRNA synthesis, or (ii) these mutations affect independently the synthesis of each viral RNA, but total effects leads similar outputs in the synthesis of all viral RNAs.

To elucidate whether these mutations affect genome replication (cRNA and vRNA synthesis) and/or transcription (viral mRNA synthesis) activities, we measured the primary transcription activity using cycloheximide (CHX), a potent inhibitor of protein synthesis (Figure 3). It is shown that CHX suppresses viral protein synthesis and thereby leads to degradation of replicated virus genome RNA but not viral mRNA since newly vRNP formation was repressed (Vreede et al., 2004; Kawaguchi et al., 2005). We utilized this method to measure the primary transcription activity that depends just only on incoming vRNP and is not affected by the replication process. In the presence of CHX, the levels of viral mRNA and vRNA were measured by qRT-PCR, and the transcription activity was represented as a ratio of viral mRNA/vRNA. This result shows that the transcription activity is not affected by these mutations, and thereby strongly suggests that these mutations affect the replication activity.

Aspartate at the position 27 is highly conserved among influenza A viruses, except for an H4N8 strain isolated from least sandpiper that contains asparagine (GenBank: ACI90144.1). We generated D27E and D27N in addition to D27V (Figure 4). The RNA levels of mRNA, cRNA, and vRNA of D27N and D27V were increased significantly compared with those of wild-type and D27E. One of possible interpretations is that uncharged amino acids at a.a. position 27 may enhance the RNA synthesis.

The mutation at the a.a. position 44 reduced the replication activity (Figure 2). To clarify the importance of the a.a. at this position, we additionally generated N44D and N44Q viruses in addition to N44I and examined the RNA synthesis activity (Figure 5). The synthesis level of each viral RNA of N44I was decreased largely, while the amounts of mRNA and cRNA of N44D and N44Q were similar to those of wild type. In addition, the amount of vRNA of N44Q was more than that of wild type. Thus, it is expected that the a.a. position 44 might be a water-soluble characteristic, and especially glutamine at this position stimulates the vRNA synthesis.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.