The high genetic diversity of influenza A virus (IAV), together with gene reassortment, forms viral quasispecies which enable IAV to acquire its efficient host adaptability. It is widely accepted that the efficient evolutionary capacity of IAV is facilitated by its highly error-prone replication process 
. Every genome copy packaged in newly produced viral progenies is synthesized exclusively by the IAV RNA polymerase (Pol), and thus it has been logically assumed that IAV Pol is a highly error-prone polymerase which provides the genomic mutations necessary for viral evolution and host adaptation. Importantly, however, the actual enzyme fidelity of IAV RNA polymerase, which is the main driving force for viral mutation synthesis, has never been characterized. On the other hand, human immunodeficiency virus type 1 (HIV-1) is also well known for its genomic hypermutability 
, and indeed HIV-1 DNA polymerase, called reverse transcriptase (RT), is the most error prone DNA polymerase among those involved in genome replication 
. However, the replication strategy of these polymerases is very different. HIV-1 RT replicates viral genomes only twice per infection cycle, and every RNA genome in the HIV-1progeny virions is actually produced by host RNA polymerase II. In contrast, IAV Pol not only is responsible for synthesizing every viral RNA genome but also amplifies viral RNAs multiple times per infection cycle, which gives IAV Pol ample chances to generate mutations in every infection event.
IAV Pol has been qualitatively studied for its RNA-dependent RNA polymerase activity using various biochemical settings 
. However, the mechanistic studies of IAV Pol complex, which determine basic kinetic parameters of the polymerases such as the nucleotide binding affinity, conformational change and chemical catalysis, have not been explored. This is because of several technical difficulties that have hindered this type of biochemical analysis for IAV Pol. First, the expression and purification of IAV Pol has been less than optimal for producing a large quantity of the active enzyme, simply because it is technically more challenging to express and purify the IAV Pol complex which consists of three different subunits, PA, PB1 and PB2, as compared to other single subunit polymerases such as HIV-1 RT. Second, polymerase studies which determine the basic kinetic parameters of DNA/RNA polymerases employ pre-steady state conditions instead of the steady state condition 
. However, since the pre-steady state condition requires the saturation of the template/primer complex used (100–200 nM) by the polymerase, a large amount of the purified polymerase is necessary for this type of analysis. Third, the IAV Pol assays use very high concentrations (mM) of primers such as ApG and cellular mRNAs, while other polymerase assays (i.e. HIV-1 RT) employ much lower concentrations of the primers (10–100 nM). Therefore, under the current IAV Pol reaction conditions available, only a fraction of the total primer is extended in each polymerase reaction, and thus the use of the 5′ end labeled primers, which are routinely used in kinetic assays of other polymerases, are not easily applicable to the IAV Pol assay.
Most frequently used polymerization assays of IAV Pol in current studies are gel-based standing start primer extension reactions that incorporate NTP substrates using ApG or cellular mRNAs as primers. In typical IAV Pol assays, one of the four NTPs is an α-32
P-labeled radioactive NTP (i.e. α-32
P-GTP), which aids in the visualization of the extended products, and the other three NTPs are nonradioactive (i.e. ATP, CTP and UTP), 
. However, while the three nonradioactive NTPs are normally used at mM ranges, the concentration of the radioactive NTP (i.e. α-32
P-GTP) is relatively very low (0.08 to 50 µM), simply because the highly concentrated radioactive NTP is not readily available, and use of a large quantity of radioactivity is technically challenging. Indeed, many recent reports that studied the steady state kinetic NTP incorporation by IAV Pol still used approximately 1/100 to 1/1000 biased nucleotide pools for the radioactive NTP (i.e. GTP) to the other three nonradioactive NTPs (i.e. ATP, CTP and UTP) 
. These unbalanced NTP pools in the reaction raise several concerns. First, it is highly likely that the use of 100- or 1000-fold lower concentration of the radioactive GTP, as compared to the three other nonradioactive NTPs, induces a serious polymerase kinetic bias: the entire IAV Pol reaction rate is simply dependent on the limited availability of radioactive GTP in the reaction, ultimately leading to the underestimation of the kinetic activity of IAV Pol. Second, biased NTP pools are a well known factor for mutagenesis 
, implying that some of the extension products, which were used for assessing polymerase activity, are actually the products of mutation synthesis. Third, it is well established that cellular NTP concentrations range between 0.25–3 mM (µM, ATP, 3,152 +/− 1,698; GTP, 468 +/− 224; UTP, 567 +/− 460 and CTP, 278 +/− 242) 
In this report, we established new assay conditions that use unbiased and biologically relevant NTP substrate concentrations and avoid the kinetic bias found in the currently available gel-based assays. Importantly, this adjustment of the assay conditions not only revealed the robust polymerase activity of IAV Pol complex, but also enabled us to investigate the enzyme fidelity of IAV Pol. Also, with this new assay condition, we were able to conduct a series of comprehensive biochemical analysis to study the impact of various reactions parameters such as temperature, metals and viral NP on the IAV Pol fidelity. Overall, our results suggest that IAV Pol is least error-prone when compared to two reverse-transcriptases, HIV-1 RT and MuLV RT and also to bacterial phage T7 RNA polymerase. Importantly, the biochemical finding of the high fidelity nature of IAV Pol introduces a new concept of the mechanistic relationship between IAV Pol fidelity and the IAV replication strategy which uniquely provides IAV Pol with ample opportunities to generate genomic mutations during the multiple rounds of cRNA and vRNA amplification in a single infection cycle.