Consecutive influenza surveillance of neuraminidase mutations and neuraminidase inhibitor resistance in Japan

Background The large consumption of neuraminidase inhibitors (NAIs) for the treatment of influenza virus infections places Japan at risk of becoming the epicenter of the global spread of NAI‐resistant viruses. Objective To clarify NA amino acid mutations of epidemic influenza viruses in Japan and their related NAI resistance. Methods A total of 1791 samples, including 396 A/H1N1pdm09, 1117 A/H3N2, and 278 B isolates, were collected to determine of their 50% inhibitory concentration (IC 50) values by NAIs (oseltamivir, zanamivir, peramivir, and laninamivir) during the Japanese seasons from 2011‐2012 to 2016‐2017. Then, 380 samples including 49 A/H1N1pdm09, 251 A/H3N2, and 80 B isolates were sequenced for the entire NA genes. Results Neuraminidase inhibitor‐resistant A/H1N1pdm09 viruses were detected at a frequency of 1.3% (5/396 isolates) in the epidemic seasons. None of the A/H3N2 and B viruses developed resistance to any of the four NAIs during the six seasons. Only five and 13 AA mutations were detected in the NA catalytic sites of A/H1N1pdm09 and A/H3N2 viruses, respectively. No mutations were observed in the catalytic sites of B viruses. Four of the five mutations in the catalytic sites of A/H1N1pdm09 consisted of H275Y, which was related to high resistance to oseltamivir and peramivir. Most (10/13) of the catalytic site mutations in A/H3N2 were associated with MDCK‐passaged induction (D151G/N). Finally, no mutations related to substantial NAI resistance were detected in the A/H3N2 and B viruses examined. Conclusion These findings suggest that the NA catalytic sites of influenza viruses are well preserved. Even in Japan, no spread of NAI‐resistant viruses has been observed, and A/H1N1pdm09 viruses carrying H275Y remain limited.

and inhaled long-acting laninamivir octanoate hydrate (Inavir ® , laninamivir), have been mainly used for the treatment of influenza virus infections. These NAIs exert their antiviral function by binding to the enzymatic catalytic sites of influenza surface protein NA. 1 In the 2007-2008, oseltamivir-resistant seasonal A/H1N1 viruses carrying NA amino acid (AA) mutation H275Y rapidly spread throughout the world. 2 We reported a prolonged duration of fever after administrating oseltamivir in patients infected with oseltamivir-resistant A/ H1N1 viruses during the 2008-2009 season in Japan, which demonstrated the association between clinical effectiveness and the acquisition of resistance to anti-influenza drugs. 3,4 In the subsequent seasons, oseltamivir-resistant A/H1N1pdm09 viruses carrying H275Y were reported, 5 causing a concern for worldwide spread. A global surveillance system has been developed to detect NAIresistant influenza isolates and their related NA AA mutations. 6 Currently, Japan is considered as the largest NAIs consuming country. 7 Thus, Japan is in an environment where the emergence and selection pressure of NAI-resistant influenza viruses are always strong compared with other countries and may become an epicenter of their global spread. Therefore, observing the transition of influenza NA AA mutations in Japan is considered important; however, comprehensive surveillance data, which include not only detecting NAI resistance and its related AA mutations but also determining NA AAs in epidemic viruses season by season, have not been fully reported in Japan. We have developed a network of physicians throughout Japan who routinely collect influenza virus samples, along with patient information. 3,4 In this study, we sequenced the full length of NA genes of epidemic influenza viruses isolated in Japan during the seasons from 2011-2012 to 2016-2017 to detect their AA mutations and examine its relationship with NAI resistance. H1N1pdm09, 251 A/H3N2, and 80 B isolates, were analyzed to obtain NA sequencing data. Nasopharyngeal swabs for influenza virus isolation were collected, and the background history of patients who had a positive result on a rapid influenza antigen test, conducted in one of the member clinics of our nation-wide study network of general practitioners, was determined. Informed consent was obtained from all patients. All study participants were outpatients, and this study did not include any patients with severe chronic respiratory diseases, renal diseases, liver diseases, or heart failure.

| Influenza virus isolation and typing
Nasal aspirates, nasopharyngeal swabs, or self-blown nasal discharge obtained from patients were soaked in virus transport medium and 75 μL of the medium was cultured using Madin-Darby canine kidney (MDCK) cells. Monolayer-cultured MDCK cells were inoculated with viruses collected from the clinical samples, and the cells were incubated at 34°C and 5% CO 2 . The wells were monitored daily until 7 days for virus growth by cytopathic effects, and the supernatant of the wells that exhibited sufficient cytopathic effects was collected. Viral RNA was extracted from infected MDCK cell culture supernatants using the Maxwell 16 LEV simply RNA Cells Kit (Promega, Madison, WI, USA). The A/ H1N1pdm09, A/H3N2, and B subtypes were determined using type-and subtype-specific primers. 13

| Measurement of virus susceptibility to NAIs
As a marker of virus susceptibility to NAIs, the IC 50 values of oseltamivir, zanamivir, peramivir, and laninamivir were determined for each influenza isolate by a fluorescence-based neuraminidase inhibition assay. 14,15 Reduced NAI susceptibility was defined according to the criteria provided by the World Health Organization. 16 The IC 50 value for each virus was compared with the median IC 50 for the same subtype viruses isolated in the same season to calculate a fold increase. For A viruses, reduced inhibition (RI) was defined as a 10-fold to 100-fold increase in IC 50 values, while highly reduced inhibition (HRI) was defined as more than 100-fold increase in IC 50 values. For B viruses, the corresponding fold increases of RI and HRI were 5-50 and >50, respectively. Viruses exhibiting RI or HRI were considered as NAI resistant.

| Sequencing
RT-PCR was performed using the A/H1N1pdm09, A/H3N2, and B RNA samples. 17,18 After amplicon preparation, deep sequencing was conducted using the Illumina MiSeq sequencing system (Illumina, San Diego, CA, USA). 19 Data processing was then performed using the pipeline prepared by Amelieff Co. (Tokyo, Japan). Finally, genome sequences were constructed using the reference sequences and filtered variants. The NA AA sequence was deduced from the obtained nucleotide sequence. In this study, a dominantly occupied nucleotide at each base position was determined as a consensus in the data processing. The NA AA lengths of A/H1N1pdm09, A/ H3N2, and B viruses were 469, 469, and 466 AAs, respectively. The Tables S1-S13.

| Nucleotide sequence accession number
The sequence data from this study were deposited into the DDBJ/ EMBL/GenBank nucleotide sequence databases under the following accession numbers: LC406948-LC407103, LC408965-LC409058, and LC409129-LC409257.

| Statistical analysis
Categorical variables were analyzed using the chi-square test. P value < 0.05 was considered to be statistically significant. All statistical analyses were performed using the jmp pro software, version 11 (SAS Institute, Inc., Cary, NC, USA).

| RE SULTS
The detection rates of circulating influenza A/H1N1pdm09, A/ H3N2, and B viruses with reduced susceptibility (RI or HRI, RI/HRI) NA AA mutation rates were compared between catalytic and non-catalytic sites of each virus (Table 2). Only five AA mutations were detected in the catalytic sites of A/H1N1pdm09 viruses, although their AA mutation rates at catalytic sites were not significantly different from those at non-catalytic sites (0.54% vs 0.45%).
For A/H3N2, only 13 AA mutations were detected in the catalytic sites. In the entire seasons examined, the AA mutation rates of A/ H3N2 viruses were significantly lower at catalytic sites than at noncatalytic sites (0.27% vs 0.65%, P = 0.001). No AA mutations were detected in the catalytic sites of B viruses, resulting in a significant lower frequency at the catalytic sites compared with those at the non-catalytic sites (0.0% vs 0.70%, P = 0.001). In the entire season examined, the AA mutations per sample at the catalytic sites of A/ H1N1pdm09, A/H3N2, and B were 0.10, 0.05, and 0.00 AAs, respectively, which confirmed the extremely low number in the AA mutations at the NA catalytic sites (Table S14).
Based on the reports on NA AA mutations associated with reduced inhibition by NAIs, 2,6,20 RI/HRI-related AA mutations were extracted from NA sequencing data examined in this study (Table 3). the non-catalytic sites was also reported to be associated with an MDCK passage. 23 These AA mutations were not detected in any of unpassaged clinical samples. 22 Finally, few AA mutations (3/4769, 0.06%, referred to  In this study, the detection rates of AA mutations in the noncatalytic sites of NA genes were also examined, and relatively many mutations to the catalytic sites were detected. In addition, the AA mutations associated with reduced inhibition by NAIs, which were indicated by the previous reports, were also included in their mutations.

| D ISCUSS I ON
However, these presumed NAI resistance-related mutations did not show any substantial reduced inhibition in this study (Table 3). These findings suggested that the direct contribution of mutations in the NA non-catalytic sites to NAI resistance was extremely rare.
RI/HRI-related AA mutations were extracted from NA sequencing data examined in this study, based on the previous reports.
The AA position numberings of A/H1N1pdm09, A/H3N2, and B viruses are based on N1, N2, and B type-specific numbering, respectively. Median IC 50 values and fold changes are shown in Tables S1-S13.
TA B L E 1 (Continued) N386S/K, which was in contrast suggested to cause a negative effect on viral fitness in the presence of the V241I and N369K mutations. 29,30 In our data (Table S15) Table 3).
The mechanism of acquiring drug resistance in NA catalytic sites for zanamivir and laninamivir is different from that for oseltamivir and peramivir, due to a difference in their side chain structures. 31 Laninamivir is licensed only in Japan and has been used increasingly in the last few years, being the main NAI used. No mutations associated with laninamivir resistance were detected in our study. This finding also suggests that the NA catalytic sites of influenza viruses are well preserved. Even in Japan, the largest consumer of NAIs, no spread of NAI-resistant viruses with mutations related to clinical ineffectiveness, such as a prolonged duration of fever, has been observed, and the emergence of A/H1N1pdm09 viruses carrying H275Y remains limited.
NAIs are currently effective for the treatment of influenza virus infections, although a continued surveillance is warranted in the future.

ACK N OWLED G EM ENTS
We thank all doctors for participating in this study and other collaborators. We also thank Hiroyuki Sasaki for his advice on this study.

CO N FLI C T O F I NTE R E S T
This investigation was partially supported by Daiichi Sankyo Co., Ltd.

H. Ikematsu has previously received honoraria from Daiichi Sankyo
Co., Ltd, for medical advice.