Baloxavir Marboxil Susceptibility of Influenza Viruses from the Asia-Pacific, 2012-2018

Baloxavir Marboxil (BXM) is an influenza polymerase inhibitor antiviral that binds to the endonuclease region in the PA subunit of influenza A and B viruses. To establish the baseline susceptibility of viruses circulating prior to licensure of BXM and to monitor for susceptibility post-BXM use, a cell culture-based focus reduction assay was developed to determine the susceptibility of 286 circulating seasonal influenza viruses, A(H1N1)pdm09, A(H3N2), B (Yamagata/Victoria) lineage viruses, including neuraminidase inhibitor (NAI) resistant viruses, to Baloxavir Acid (BXA), the active metabolic form of BXM. BXA was effective against all influenza subtypes tested with mean EC50 values (minimum-maximum) of 0.7 ± 0.5 nM (0.1–2.1 nM), 1.2 ± 0.6 nM (0.1-2.4), 7.2 ± 3.5 nM (0.7–14.8), and 5.8 ± 4.5 nM (1.8–15.5) obtained for A(H1N1)pdm09, A(H3N2), B(Victoria lineage), and B(Yamagata lineage) influenza viruses, respectively. Using reverse genetics, amino acid substitutions known to alter BXA susceptibility were introduced into the PA protein resulting in EC50 fold change increases that ranged from 2 to 65. Our study demonstrates that currently circulating viruses are susceptible to BXA and that the newly developed focus reduction assay is well suited to susceptibility monitoring in reference laboratories.


Introduction
Despite increased vaccination rates in many parts of the world, influenza continues to cause high levels of morbidity and mortality in high-risk groups [1], particularly when influenza seasons are dominated by A(H3N2) viruses [2]. Influenza antivirals are available for the short-term prophylaxis of individuals to prevent influenza infection, but their primary use has been to treat severely ill patients, many of whom are hospitalised. Two classes of influenza antivirals have been licensed for many years, the M2 ion channel inhibitors and neuraminidase inhibitors (NAIs). Clinical use of the M2 ion channel inhibitors, amantadine and rimantadine, is limited as close to 100% of circulating influenza A viruses contain an amino acid (AA) substitution at residue 31 of the M2 protein (S31N) that confers resistance to these compounds [3]. Four NAIs are licensed in different parts of the world, of which oseltamivir is the most widely available and commonly used. Oseltamivir resistance has been widespread amongst certain groups of viruses in different periods of time (e.g. seasonal H1N1 between 2007 and 2009 [4,5]), but for the last seven years the frequency of viruses that circulate with reduced NAI susceptibility has remained at less than 5% [6]. The licensure of alternative antivirals, especially those with different modes of action to NAIs, is likely to be of benefit if oseltamivir resistant viruses emerge. In addition, combination therapy may be a strategy to improve clinical effectiveness compared to NAI monotherapy [7].
Baloxavir marboxil (S-033188, BXM) is an influenza polymerase inhibitor that was licensed for the treatment of uncomplicated influenza in Japan and the US in 2018 [8]. BXM is a prodrug that is hydrolysed by the enzyme arylacetamide deacetylase to the active form baloxavir acid (S-033447, BXA) [9]. BXA is a small molecule inhibitor of the highly conserved cap-dependant endonuclease (PAN) of influenza A and B viruses [10]. Inhibition of the PAN disrupts endonuclease function and as a consequence the cap-snatching mechanism of the influenza polymerase [10]. Treatment of uncomplicated influenza with BXM as a single oral dose was shown in a Phase III clinical trial to reduce influenza symptom duration by 26.5 hours compared to placebo, similar to the reduction time achieved with oseltamivir [11]. However, 24 hours post-drug administration the reduction in viral load in BXM treated patients was twice as large as oseltamivir treated patients [11].
In post-treatment samples obtained from phase II and III clinical trials of BXM, A(H3N2) and A(H1N1)pdm09 viruses with reduced BXA susceptibility were detected and shown to carry AA substitutions at position 38 of the PAN including I38T, I38M and I38F [12]. In the phase II clinical trial (which involved predominantly A(H1N1)pdm09 viruses) a substitution at I38 emerged in 4 of 182 patients (2.2%), while the phase III study (which involved predominantly A(H3N2) viruses) showed a frequency of an I38 variant in 36 of 370 (9.7%) of BXM recipients [11]. The highest frequency of viruses with reduced BXM susceptibility has been reported from a paediatric study, where 18 of 77 (23.4%) of patients had viruses bearing substitutions at I38 following treatment [12]. The transmissibility of the PAN /I38 variants between patients in the absence of drug treatment is currently unknown, but in vitro studies suggest that these viruses do not replicate as well as the equivalent wildtype strains [10].
Given the frequency of viruses detected following BXM treatment with reduced BXA susceptibility, and the potential for widespread use of the drug in clinical practice, it is important to conduct surveillance of circulating strains for BXM susceptibility. This study aimed to develop a highthroughput and reproducible phenotypic assay for surveillance purposes to determine the BXA susceptibility of recently circulating influenza viruses in the Asia-Pacific region, thereby providing baseline susceptibility data to which prospective samples can be compared to in the future.

Site Directed Mutagenesis and the Generation of Recombinant Viruses
The reverse genetics plasmid sequencing was used to confirm AA changes in each plasmid. An eight-plasmid reverse genetics method adapted from Hoffman et al was used to generate recombinant viruses [13]. Alterations to the methods include, HEK-293T and MDCK-TMPRSS2 were seeded at an 8: 1 ratio with a total of 5x10 5 cells and COS-7 and MDCK-TMPRSS2 at a 3:1 ratio with a total of 3x10 5 cells, for influenza A and influenza B virus experiments, respectively. GeneJuice® Transfection reagent (Merck Millipore, USA) was used for transfection. MDCK-TMPRSS2 were infected 72 hours post-transfection using 1 mL of supernatant from the co-culture. Virus growth was determined using haemagglutination assay and 1% (v/v) turkey red blood cells.

BXA Cytotoxicity Assay
The cytotoxicity of BXA was measured to identify non-toxic drug concentrations suitable for use in vitro. The inner wells of 96 well plates (Corning, USA) were seeded with MDCK-SIAT cells at a concentration of 2.5 x 10 5 cells/mL (100 μl/well) and incubated overnight at 37°C in a 5% CO2 gassed incubator. The BXA concentration range of 50 μM to 0.4 μM was obtained from a serial two-fold dilution of BXA in DMSO (Sigma Aldrich, US) and further diluted in the final MM supplemented with 2 μg/mL TPCK-trypsin. One well was left free of BXA as a negative control. Treated cells were incubated at 35°C in a 5% CO2 gassed incubator for 24, 48 or 72 hours. Cell viability was determined using the CellTiter-Glo® Luminescent Cell Viability Assay as per manufacturer's instructions (Promega, USA) and luminescence was measured using a FLUOstar Optima luminometer (BMG Labtech, Germany). The BXA concentration that reduced cell viability by 50% compared to the cell only control (CC50) was calculated using non-linear regression analysis (GraphPad Prism, USA).

Virus Titration
Virus titration is required to select a suitable virus dilution for the focus reduction assay. MDCK-SIAT cells were seeded in the inner wells of 96 well plates (Corning, USA) at a concentration of 2.5 x 10 5 cells/mL (100 μl/well) and incubated overnight at 37°C in a 5% CO2 gassed incubator. The experiment was only continued if the cell monolayer was 100% confluent the following day. MDCK-SIAT cells were infected and immunostained with previously described methods [14]. in 2% skim milk. Plates were then washed and incubated for one hour with goat anti-mouse IgGhorse radish peroxidase (Biorad, US) secondary antibody, diluted 1:1000 in 2% skim milk. Plates were again washed and then incubated for ten minutes in the dark with TrueBlue™ Peroxidase Substrate (KPL, US) followed by three washes with distilled water, the water was then removed, and plates were allowed to dry. Focus forming units (FFU) were quantified using the Immunospot BioSpot 5.1.36 (CenturyLink Inc, US).

BXA Focus Reduction Assay (FRA)
The concentration of BXA required for a 50% reduction in FFU (EC50) was used to determine susceptibility of influenza viruses to BXA. MDCK-SIATs were seeded and infected as described in

X= Mean FFU
Using the percentage inhibition, the EC50 for BXA of each virus was determined using non-linear regression analysis (GraphPad Prism, USA).

Statistical Analysis
The Linear regression analysis and unpaired student's t-tests were performed using GraphPad Prism (USA) where p-values < 0.05 were considered statistically significant. To evaluate assay reproducibility, FRA was performed with replicate (n = 48) wells of positive and negative controls (± virus) on a 96-well plate and Z factors were calculated using the equation outlined in [15].

Cytotoxicity of BXA
To determine the maximum working drug concentration for use in in vitro assays, a CellTiterGlo assay was used to measure cell viability in the presence of increasing concentrations of BXA in MDCK-SIAT cells at 24, 48 and 72 hours. The MDCK-SIAT cell cytotoxicity (CC50 -the 50% reduction of cell cytotoxicity compared to a cell only control) of BXA was 34.1 ± 1.9 μM, 10.1 ± 2.1 μM and 7.8 ± 0.9 μM at 24, 48 and 72 hours, respectively.

Reproducibility of BXA Focus Reduction Assay
Several factors were tested to ensure the reproducibility of the FRA method and it suitability for use as a robust, high throughput screening assay. Z scores were determined at 18  This difference was statistically significant (P<0.0001) based on an unpaired student's T test.

Susceptibility of NAI Resistant Viruses to BXA
Eleven viruses with neuraminidase AA substitutions that confer reduced susceptibility to oseltamivir, zanamivir, peramivir or laninamivir were screened for susceptibility to BXA using the FRA (Table 1).
All viruses tested had EC50 values within the expected range for influenza A (0.1 to 2.4 nM) and influenza B (0.7 to 15.5 nM). These data demonstrate that BXA is active against influenza strains which have reduced susceptibility to NAIs.

Susceptibility and Phenotypic Screening of PAN I38 Amino Acid Substitution to BXA
Substitutions in the PAN /I38T/I38M/I38F have been identified in some influenza viruses in BXM clinical trials from patients following treatment [12]. Therefore we evaluated the BXA susceptibility of viruses with these substitutions in the FRA.

Discussion
This study aimed to establish a robust and reproducible assay to determine the susceptibility of influenza viruses circulating in the past seven years to the PAN inhibitor drug BXA. In addition, we   [12], and it is likely that additional sites will be detected as clinical use of the drug increases.
This study provides information on the baseline susceptibility of a large number of recently circulating influenza viruses across all relevant subtypes and lineages in the Asia-Pacific region. It will be important to continue to test circulating viruses for BXM susceptibility as the antiviral continues to be licensed and used more widely to better understand the molecular determinants of BXA susceptibility, the frequency with which viruses with reduced susceptibilty occur, and whether they have the capacity to transmit within the community in the absence of drug selective pressure.