The molecular-genetic and phylogenetic analysis of the hemaglutinin gene of influenza viruses

Aim. To perform a phylogenetic and molecular-genetic analysis of the HA genes of influenza viruses that circulated in Ukraine during the 2016–2017 epidemic season, and to compare them with those that circulated in the world. Methods. Samples (nasopharyngeal swabs from patients) were analyzed using the real-time polymerase chain reaction (RT-PCR). Phylogenetic trees were constructed using MEGA 7 software. 3D structures were constructed in Chimera 1.11.2rc software. Results. Ukrainian isolates from the 2016–2017 season have substitutions in the antigenic sites which were not detected earlier; they and can influence the antigenic properties of viruses. Otherwise the A(H3N2) and B/Victoria viruses retained the similarity to the vaccine strains. For the A(H1N1)pdm09, a higher similarity to the vaccine strain recommended for the 2017–2018 epidemic season was observed. Conclusions. In the 2016‒2017 epidemic season, all influenza viruses — A(H3N2), A(H1N1)pdm09 and B/Victoria acquired a number of unique amino-acid substitutions in the HA gene. The results of this study reaffirm the continuous genetic variability of circulating seasonal influenza viruses and the need for continued system-atic antigenic and molecular surveillance.


Introduction
Influenza viruses cause global epidemic infections each year, a peak is from December to March. These pathogens have also contributed to six global pandemics known so far. Influenza viruses are a group of pathogens in the family Orthomyoviridae, which are classified into 6 genera (A,B,C, Thogotovirus, Isavirus and new unnamed genus). Type A influenza viruses are categorized based on their surface glycoproteins: hemagglutinin (HA) and neur- ISSN 1993-6842 (on- aminidase (NA). So far 17 HA and 9 NA subtypes have been identified [1]. Hemagglutinin is known to be a major target region of neutralizing antibodies which inhibit binding with sialic acid receptors effectively. Mutations in the HA antigenic sites designated A, B, C, D, and E in IAV of H3N2 subtype, may result in strains which can escape recognition by preexisting neutralizing antibodies [2]. It is known that the H1 HA molecules have four distinct antigenic sites: Sa, Sb, Ca, and Cb [3]. Influenza viruses B/Victoria have four antigenic sites -120-loop, 150-loop, 160-loop and 190-helix [5]. As a result, these sites consist of the most variable amino acids in the HA molecule of the seasonal human H1N1 viruses that have been subjected to antibodymediated immune pressure (antigenic sites H1). Gradual accumulation of mutations at these sites is noted to be an integral component of evolutionary dynamics and impacts viral survival and fitness. This evolutionary mechanism, known as antigenic drift, is the basis for frequent updating of the composition of seasonal influenza vaccines [4]. The World Health Organization (WHO) encourages National Influenza Centers (NICs) to conduct ongoing influenza virologic surveillance, to monitor spread of viruses and their continuous evolution. Combining data from phylogenetic and molecular analyses of influenza viruses is essential to detect virus variants that have undergone antigenic drift, variants with enhanced virulence or variants with reduced sensitivity to antivirals. Such combined genetic, antigenic and phylogenetic analyses provide improvements in the process of vaccine virus selection and inform patient treatment regimens [5].

Methods
Nasal-throat swabs were taken from the patients with suspected influenza and SARI (severe acute respiratory infections). Samples were analyzed using a real-time polymerase chain reaction (RT-PCR). Influenza viruses were isolated in MDCK and MDCK-SIAT cell culture. The isolates were later used for the strain identification and sequencing. The sequencing of influenza viruses isolates was performed in the World Influenza Center in London using the technology of RNA-SEQ. The sequences of influenza viruses from other countries were received from web-site GISAID (the Global Initiative on Sharing All Influenza Data -http://platform.gisaid.org) using BLAST analysis. The sequences were aligned using ClustalW algorithm. Phylogenetic trees were built by the neighbor joining method applying Kimura 2-parameter model. The nucleotide sequences were translated into the amino acid sequences using MEGA 7 software [6]. 3D structures were constructed in Chimera 1.11.2rc software (https://www.cgl.ucsf.edu/ chimera/).

Results and Discussion
Influenza viruses type A: A(H1N1)pdm09 and A(H3N2); and influenza viruses type B, lineage B/Victoria were circulated during 2016-2017 epidemic season.
The epidemic season of 2016-2017 years was less intensive than the previous epidemic season of 2015-2016 years. In Ukraine A(H3N2) viruses played a main role in this season, as well as in other European countries. According to the Ministry of Health of Ukraine, in the 2016-2017 season 30 fatal cases of influenza were registered, while in the previous season this figure was 10 times higher [7]. In total in the 2016-2017season, according to the official statistics, only 106,753 persons were vaccinated against the flu, 15 % less than in the previous season. HA gene sequences were recovered from GISAID and phylogenetic trees were constructed to determine the genetic relationships of Ukrainian isolates with the reference viruses and viruses circulating in other countries in the same period. Complete HA amino acid sequences of Ukrainian influenza isolates were compared to those of the vaccine viruses to identify substitutions that might impact the vaccine effectiveness.

Characterization of mutations of influenza viruses type A(H3N2)
The HA genes fell within genetic group 3C. This group has three subdivisions: 3C.1, 3C.2 and 3C.3; subdivisions 3C.2 and 3C.3 have undergone further evolution to yield subclades designated as 3C.2a, 3C.3a and 3C.3b respectively. New subclade 3C.2a1 was emerged in the last 2016-2017 epidemic season. All viruses belonged to the genetic group 3C.2a1 had acquired amino acid substitutions N171K, I406V and G484E (I77V and G155E in НА2) [7] (Fig. 1). The majority of Ukrainian isolates belonged to a new genetic group 3C.2a1. Only three isolates belonged to the genetic group 3C.2a. The mutation N128T resulting in gain of glycosylation site was in the antigenic site B. This is one of the options for defending the virus from immune protection of the human body. Some of mutations can influence the antigenic properties of viruses as they are found in the antigenic sites. The mutation in 144 position is located in the antigenic site A, two mutations 159 and 160 -in the antigenic site B, position 311 -in the antigenic site C [3]. None of the substitutions in HA relate to known MDCK culture-induced substitutions [8].

Characterization of mutations of influenza viruses type B, lineage B/Victoria
Phylogenetic comparison of the influenza virus type B HA genes has shown that all investigated isolates are genetically related to the vaccines strain B/Brisbane/60/2008 (Fig. 2).
Currently among the viruses of influenza В the genetic branch В / Victoria the strains belonging to the genetic group 1А are circulating in the world [9]. All these isolates have specific mutations, which attached these isolates to the 1A group. On the phylogenetic tree of the gene of HA, two reference isolates belong to the genetic group 1B, but now these isolates are not circulating in the world.
The comparison of molecular-genetic changes was carried out with the reference strain B/Malaysia/2506/2004. All the viruses had mutations N75K, N165K, S172P, and I199T comparing with the reference strain (Fig.2). The mutations N165K and I119T are placed in the antigenic site of 160-loops and 190-helix respectively. The accumulation of such mutations in the antigenic sites leads to the changes in the antigenic properties of viruses [5].
Several mutations were in the antigenic sites: I117V and N129D -in 120-loop antigenic site, V146I -in 150-loop antigenic site. Ukrainian isolates were similar to the viruses from different parts of the World. Despite the mutations, the viruses retained genetical similarity to the vaccine strain B/Brisbane/60/2008 at 99 % and strain B/Brisbane/60/2008 was recommended for the vaccine composition for the next 2017-2018 epidemic season [9].

Characterization of mutations of influenza viruses type A(H1N1)pdm09
The percentage of pandemic influenza viruses was the smallest in the epidemic 2016-2017 season. Only two A(H1N1) viruses were isolated in Ukraine -from Kharkiv and Odesa. The viruses isolated in Ukraine in the 2015-2016 season belong to the 6B genetic group, which has two new subgroups 6B.1 and 6B.2 that emerged this season. All viruses belonged to the genetic group 6B.1 in the 2016-2017 season. The amino acid sequence analysis was performed in comparison with the vaccine strain A/California/07/2009. This strain was in the vaccine composition since 2009, and only according to the recommendation for the 2017-2018 epidemic season it was replaced by a new one [9]. Ukrainian viruses isolated in the 2016-2017 epidemic season belonged to new genetic group 6B.1 and were related to the viruses from Maldives and England. It may indicate the place of origin of Ukrainian viruses. The S162N main substitution appeared in the Sa antigenic site and was observed in all isolates of the 6B.1 genetic group, and this mutation influenced the antigenic properties of viruses [7]. That is why in the recommendation for the 2017-2018 epidemic season the vaccine strain was replaced by a new one -A/ Michigan/45/2015, which belongs to the new genetic subgroup 6B.1 [9].
All viruses had mutations -D97N, S185T, and S124N (HA2), K163Q, A256T, K283E. The mutation in position 185 was in the antigenic site Sb. The S185T substitution falls within a domain defining the receptor binding site (RBS). It has been reported that substitutions in or near the RBS can influence the antigenic properties of A(H1N1)pdm09 viruses [10], and attachment of oligosaccharide chains to the antigenic sites has been suggested to facilitate the immune evasion. None of these substitutions were associated with known adaptation to propagation in MDCK culture [11]. The