Experimental Infection of Adapted Influenza B Virus in Mice Model

Over the years influenza B virus (IBV) contribute annual disease and can lead to serious respiratory disease among humans. More attention should be paid to the mammalian adaptive processes of B viruses and development of vaccines against current influenza. Because of preclinical trials of anti-influenza drugs are conducted mainly on mice, we developed adequate animal model using antigenically-relevant IBV strain for testing anti-influenza drugs and protective efficacy of flu vaccines. We serially passaged Victoria lineage (clade 1A) IBV 17 times in BALB/c mice. The adaptive amino acid substitutions were found in HA (T214I) and NA (D432N). By the electron microscopic examination, we showed spherical and elliptical shapes of IBV. Light microscopy showed that mouse-adapted B virus caused influenza pneumonia on day 6 post inoculation. We evaluated the illness pathogenicity, viral load and histopathological features of mouse-adapted IBV and estimated anti-influenza drugs and vaccine efficiency in vitro and in vivo. Assessment of investigational anti-influenza drug oseltamivir ethoxisuccinate and flu vaccine Ultrix® revealed effectivity against our mouse-adapted influenza B virus.


Introduction
Influenza B virus (IBV) belongs to the family Orthomyxoviridae [1]. They were first isolated in 1940, and since the 1980s two IBV genetic lineages have been identified: B/Victoria/2/87 (B/Vic) and B/Yamagata/16/88 (B/Yam). These lineages differentiated by differences in hemagglutinin (HA) and neuraminidase (NA), which almost have no antigenic crossover in the hemagglutination inhibition assay [2,3]. Due to the fact that IBV has been isolated from both humans and seals, the reservoir remains unknown [4,5]. Infection of human caused by IBV can lead to serious respiratory disease, complications of which are particularly common among children of primary school age (5-8 years) [6]. Data for the United States for each epidemic season from 2004-2011 (excluding the 2009 pandemic) show that between 22% and 44% of all childhood influenza-related deaths were caused by IBV infection. More over from 2004 to 2013 Canadian researchers found significantly higher mortality rates due to IBV compared to influenza A virus in children younger than 16 years of age [7]. In Europe influenza B accounted for 63% of all influenza cases in the 2017-2018 epidemic season [8]. A number of studies using ex-vivo (explant) cultures of human bronchus or lung have shown that IBV are capable of causing severe lower respiratory tract infections; these frequently lead to fatal complications [9].
Seasonal influenza vaccines are divided into types: trivalent, which consists of influenza A/H1N1, A/H3N2, and one influenza B strain (B/Yam or B/Vic); or quadrivalent, which contains all four strains. The emerging threat of IBV has been recently recognized [10], and seasonal influenza vaccines are moving towards quadrivalent types [11,12,13]. Vaccines designs seek to provide protection against seasonal influenza viruses by eliciting antibody responses to surface viral HA proteins. Constant antigenic drift in HA necessitates regular updating of vaccine strains to ensure that the antigenic profile of circulating strains and vaccine components match [14]. According to the CDC, the effectiveness of quadrivalent vaccine is 28% among the especially susceptible, namely children within the age group of 9-17 years of age [15].
Despite the fact that IBVs has repeatedly caused human epidemics, its genetic determinants of virulence and transmission are still poorly understood. Limited data on the range of hosts and the absence of an animal model complicate several areas: study of pathogenicity factors; IBV modes of transmission; and assessment of antiviral drugs and vaccines effectiveness. The aim of this study was to develop better infection models, using clinically-relevant viruses, which facilitate testing of (anti-influenza) vaccineinduced protection. BALB/c mice were infected with mouse-adapted influenza B virus and were characterized for illness, inflammation, viral load and histopathology; and also, were estimated antiinfluenza drugs and vaccine efficiency in vitro and in vivo.

Materials and Methods
All manipulations with animals were approved by the Ethics Committee of the Federal Research Center of Fundamental and Translational Medicine (No. 2017-15).

Viral adaptation
The virulence of the B virus was increased by serial passages in the lungs of 8-week-old male BALB/c (n, 7 per group) mice (State Research Center of Virology and Biotechnology VECTOR (FSRI SRC VB VECTOR), Novosibirsk, Russia). Seven mice were lightly anesthetized with Rometar (20 mg/kg) (Bioveta, Czech Republic) and intranasally infected (i.i.) with 50 µl of phosphate-buffered saline (PBS) containing 10 4 TCID50/ml (50% tissue culture infective dose) of the wild type IBV strain В/Novosibirsk/40/2017 (В/2017) that were isolated from human in 2017 in Novosibirsk, Russia. Three of seven mice from each passage with the most evident symptoms were sacrificed by decapitation on 3 rd day post-infection (d.p.i.). The lungs of these mice were used to prepare 10% homogenates in PBS. Subsequently, new groups of mice were anesthetized with same anaesthetic and i.i. with 50 µl of 10% lung homogenate. In parallel, viral replication of viruses in the lungs of the sacrificed mice was measured by titration of a 10% homogenate using MDCK cell culture [16]. Four of seven mice from each group, at each passage, were monitored daily for 14 d.p.i for signs of illness, weight loss, or lethality. After 17 passages in total were registered the clinical signs: significant reduction in body weight (up to 30%); hypothermia; ruffled fur; mice began to huddled together. The wild type IBV strain В/2017 and mouse-adapted variant (strain В/Novosibirsk/40/2017-MA (B/2017-MA)) were patented [17]. The mouse infectious dose (MID50) of the virus B/2017-MA was 4.6±0.26 log10TCID50/ml, or 1.88 TCID50.
To evaluate the pathogenicity of B/2017 and B/2017-MA viruses, groups of six 6-week-old male BALB/c mice (n, 10 per group) were anesthetized with Rometar (20 mg/kg) and i.i. with 50µL of PBS containing 10 4 TCID50/ml and 10 MID50, respectively. Intact mice (n, 3 per group) were i.i. with 50 µl of PBS (pH 7.2) and serve as control. Body weight and temperatures changes, and mouse survival rate were monitored daily for 14 d.p.i., and mice that lost more than 25% of their body weight were euthanized. Body weight was measured by using a laboratory animal weighing analytical balances MASSA-К VК-1500 (MASSA-К, Russia), and body surface temperature was taken from the ear canal using a hand-held infrared thermometer «AccuVET» (Mesure technology Co., LTD).
To detect the tissue distribution of B/2017 and B/2017-MA viruses, on days 3 and 6 p.i., three mice were sacrificed and organ samples of lungs, brain, heart, liver, kidneys, spleen harvested in 1 mL of PBS. Then samples were homogenized and centrifuged, viral titers in the homogenized supernatants were determined by the Kerber method with Ashmarin-Vorobyov modification. To assess by light and electron microscopy pathological lesions from mice infected with of B/2017 and B/2017-MA viruses, their lungs were harvested at 3 rd and 6 th d.p.i.

Light microscopic examination
Lungs from 3 animals in each group (`В/2017 infected` and `В/2017-МА infected`) were taken for examination by light microscopy on the 3 rd and 6 th d.p.i. and subsequently fixed in 4% formalin solution, dehydrated (according to the standard procedure), and embedded into paraffin. Then 4-5 microns thick paraffin sections were obtained using a HM 340E rotary microtome (Carl Zeiss, Germany) and stained by H&E method. Light microscopy and photography were carried out using an Axioskop 40 microscope (Carl Zeiss, Germany).

Electron microscopic examination
Lung samples were taken on the 3 rd and 6 th d.p.i. with В/2017 and В/2017-МА viruses. Samples were fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer pH 7.4 for 4 hours at 4℃; re-fixed with 1% osmium tetroxide in 0.1 M phosphate buffer pH 7.4 at 4 °C for 2 h; then dehydrated in ethanol (50°, 70°, 96°, 100°) followed by acetone and Araldite-Epon mixture (1:6) (SPI, USA) with the addition of the catalyst DMP-30 and polymerized at 60℃. Semi-thin sections were prepared from solid blocks, stained with Azur II and examined in a light microscope to highlight areas for ultrathin section. Ultrathin sections were cut by EM UC7 ultramicrotome (Leica, Wien, Austria). Sections were stained with uranyl acetate followed by lead citrate (SPI, USA). The samples were examined on a transmission electron microscope LIBRA 120 (Carl Zeiss, Germany) at 100 kV. The images were made by a digital camera Veleta (SIS, Germany).

Sequencing and GISAID accession numbers
Viral RNA was extracted using QIAamp Viral RNA Mini Kit (QIAGEN) according to the manufacturer's instructions. Whole genome amplification of influenza B genome was performed using SuperScriptTM III One-Step RT-PCR System with PlatinumTM Taq High Fidelity DNA Polymerase (Thermo Fisher Scientific) with modifications [18].
Products of PCR were analyzed by agarose gel electrophoresis, and sequencing was performed using the Illumina MiSeq platform. Paired-end libraries for MiSeq platform were prepared using Nextera XT DNA Library Prep Kit (Illumina). The sequencing library was quantified using NEBNext Library Quant Kit (NEB). Library size was assessed using Agilent Bioanalyzer 2100.

Genetic analysis
The IBV nucleotide sequences being investigated were combined with sequences retrieved from the GISAID database. For multiple alignments, a MUSCLE programme was used [19]. Comparative pairwise sequence alignment of 2 investigated strains was performed via BioEdit. Phylogenetic trees were built via MEGA 5 using Maximum Likelihood, utilizing the general time reversible (GTR) nucleotide substitution model. Bootstrap support values were generated using 500 rapid bootstrap replicates.

Determination of susceptibility to neuraminidase inhibitors
The susceptibility of B/2017 and B/2017-MA strains to oseltamivir (Hoffmann-La Roche, Basel, Switzerland) were evaluated by published NA inhibition assays [20,21]. Briefly, viruses were standardized to a NA activity level 10-fold higher than that of the background, as measured by the production of fluorescent product from methylumbelliferyl-N-acetylneuraminic acid (MUNANA) substrate (Sigma-Aldrich, Darmstadt, Germany). Drug susceptibility profiles were determined by the extent of NA inhibition after incubation with 3-fold serial dilutions of NAIs. The 50% inhibitory concentrations (IC50) were determined from the dose-response curve.

Determination of anti-influenza drugs efficacy
We studied anti-influenza efficacy of oseltamivir etoxisuccinate and Tamiflu® on 6-8-week-old BALB/c mice (FSRI SRC VB VECTOR) (n, 10 per group). All mice of groups №№1-3 were lightly anesthetized with Rometar (2 mg/kg) and then i.i. with 10 MID50 of B/2017-MA strains in 50 µl. Mice of group №4 were i.i. with 50 µl of PBS and served as control. Then mice of group №1 were treated per os by 25 mg/kg/day of oseltamivir etoxisuccinate with dose of (200 µl of each) during 5 d.p.i.; mice of group №2 were treated per os by same dose with Tamiflu® during 5 d.p.i. Mice from group №3 received 200-µl of distilled water per os during 5 d.p.i. All animals were monitored for illness signs, weight loss, temperature changes, mortality, and lethality over 14 d.p.i.
On the 14 th day after second immunization mice were lightly anesthetized with Rometar (20 mg/kg) (Bioveta, Czech Republic) and i.i. with 50 µl of PBS containing 10 4 TCID50 of the В/2017-МА virus or sterile PBS. At 3 rd and 6 th d.p.i., 3 animals from each group were humanely euthanized for tissue collection. Lungs were collected for virus titer quantification, light and electron microscopic examinations. At 21 d.p.i., mice were bled from the submandibular vein for serology. Clinical signs of disease such as body weight and temperature changes, mortality and morbidity were monitored daily throughout the study.
For multiple comparisons, two-way analysis of variance (ANOVA) was performed. A P value below 0.05 was considered significant.

Viral adaptation
To study the adaptation of B virus isolated from human, we serially passaged the wild type IBV Victoria lineage strain В/2017 in BALB/c mice. After 17 passages total, mouse-adapted B virus В/2017-MA was obtained. We compared the pathogenicity of В/2017 and В/2017-MA strains. Groups of twelve 6-8-week-old male BALB/c mice were i.i. with 50µL of В/2017, or В/2017-MA viruses at 10 4 TCID50 (10 MID50). Body weight and temperature changes, morbidity and mortality were monitored during 14 d.p.i, and the PBS-inoculated group of mice served as the control. On the 3 rd and 6 th d.p.i., 3 animals from each group were sacrificed by decapitation, and internal organs samples (lungs, brain, heart, liver, kidneys, spleen) were taken for comparative virological analysis. On day 3 and 6 p.i., lungs were taken for examination by electron microscopy.
All mice i.i. with В/2017 virus survived and showed no obvious clinical signs such as body weight and temperature changes ( Figure 1A,B). In contrast, in the another experimental group of mice i.i. with В/2017-MA were ditected gradually weight loss: approximately 10% at 3 d.p.i., 15% at 4 d.p.i., 20% at 5 d.p.i., 25% at 6 d.p.i., and 30% at 7 d.p.i. (Figure 1A). Temperature measurements indicated the peak infection time frame to be from the 4 th to the 7 th d.p.i. (Figure 1B).  At 3 and 6 dpi, three mice were sacrificed and their organs (lungs, brain, heart, liver, kidneys, spleen) were harvested. Viral titers in the collected organs were determined by the Kerber method with Ashmarin-Vorobyov modification. The strain B/2017 could replicate in mouse lungs only 3 days with titers 2.9±0.2 log10TCID50/ml. In contrast, strain B/2017-MA replicated very well in mouse lungs, with higher titers and longer period: on 3 rd d.p.i. titer was 3.5±0.1 log10TCID50/ml, and on 6 th d.p.i. -3.3±0.3 log10TCID50/ml. Therefore, the mouse-adapted B virus strain B/2017-MA displayed much higher replication capability in mouse lungs.
Histopathological analysis of mice lungs infected with the B/2017 virus were found out a slight damages such as small number of eosinophilic cells in the bronchioles, small blood filling of the capillaries, and edema with a high protein content on 6 th d.p.i. (Figure 1C). In contrast, on the same d.p.i., pathomorphological changes in lungs of mice infected with the B/2017-MA virus were more pronounced due to viral triggered apoptosis, leading to desquamation of bronchiol epithelium. Also were seen greater number of eosinophilic cells in the bronchioles, lymphocytic infiltration of lungs various parts, and capillaries stasis ( Figure 1C). All of the pathomorphological changes listed were predominantly located in the root, cranioventral and middle regions of the left and right lungs of infected B/2017-MA virus mice.
Electron microscopic examination revealed the budding of virions from the surface of type 1 alveolar cells on the 3 rd d.p.i. in samples from B/2017-MA group mice ( Figure 1D,E). Interestingly to note that various virion morphologies were seen: spherical or elliptical, but no filamentous.

Sequencing and Genetic analysis
According to nucleotide sequences analysis of all eight genome segments, the B/2017 virus and, consequently, its mouse-adapted variant B/2017-MA belong to the B/Vic genetic lineage. In addition, analysis of these strains` HA amino acid sequences revealed mutations (I117D, N129D, V146I) relative to earlier reference strains. These mutations are characteristic of strains belonging to the 1A genetic subgroup of the B/Vic lineage. The strains` NA substitutions were also found feature amino acid substitutions characteristic of genetic group 1A of the B/Vic lineage (N340D, E358K, S295R, I120V, and K220N). To assess the phylogenetic relationships between B/2017 and B/2017-MA, all genome segments were analyzed using phylogenetic dendrograms. Analysis used IBV nucleotide sequences, available in the GISAID database, isolated from residents of Russia and Kazakhstan, as well as vaccine and reference strains according to WHO classification.
According to dendrograms (Suplementary Materials, Figures 1-8), the studied strains form a common phylogenetic group with other isolates from Novosibirsk, as well as strains from the neighbour Altai region and the Republic of Kazakhstan. At the same time, all of them are phylogenetically distanced (although only slightly) from IBV strains isolated in other Russian regions.
To identify strains that are most genetically related to the B/2017 and B/2017-MA strains, BLAST analysis was performed. The results revealed that studied strains are 99-100% identical to the IBV variants that circulated in the human population in the Novosibirsk region, Altai, and Republic of Kazakhstan (Suplementary Materials, Table 1). Thus, strain B/2017 is typical genetic variant of the IBV that circulated during 2016-2017 epidemic season, and is the most genetically related to the strains that circulated in Asia at that time.
Comparative analysis of nucleotide and amino acid substitutions between the two strains (B/2017 vs B/2017-MA) showed the presence of synonymous (not leading to amino acid substitution) nucleotide substitution in the PB1 segment -A2175G. Additionaly, nucleotide substitution in the HA segment (C641T) led to the amino acid substitution in the HA protein (T214I). According to the Flusurver resource [24], the detected substitution is rare and present in 0.46% HA of IBVs isolated from 2008 to 2016. The identified amino acid substitution is localized in the antigenically active subunit HA-HA1, which can potentially affect the biological properties of the virus. In the sequence encoding , a nucleotide substitution (G1294A), which leads to the amino acid change D432N, was detected. According to the Flusurver resource [24], this substitution has only been found in one strain (B/Hawaii/37/2017) so far.

Assessment of Antiviral Drug Therapy In Vitro and In Vivo
In our work we estimate the IC50 of oseltamivir ethoxisuccinate that necessary to reduce NA activity of Also in our study were performed anti-influenza drugs efficiency in vivo, which showed no significant body wight and temperature changes between groups of animals i.i. with B/2017-MA strain and then treated per os during 5 days with oseltamivir ethoxisuccinate or Tamiflu® ( Figure 1F,G). All treated with antiinfluenza drugs mice lost no more than 10% of body weight and began to recovere from the 7 th d.p.i. Also no hipothermia were detected among them. Therefore, was shown that innovative drug (oseltamivir ethoxisuccinate) revealed high effectiveness against the mouse-adapted B virus similarly Tamiflu®.

Assessment of Vaccine Efficiency Against Mouse-adapted Influenza B Virus In Vivo
We performed assessment of vaccine efficiency against mouse-adapted B/2017-MA strain in vivo.
Previously BALB/c mice were immunized with purified surface antigens from influenza strain Infection induced by B/2017-MA strain in non-immunized mice was characterized by body weight lost and hypothermia from the 2 nd d.p.i, and conjunctivitis (up to 30%), with onsets between days 1 to 3 d.p.i. (Figure 2A,B,C). All mice of this group huddled together, and their fur was ruffled оn the 4 th d.p.i. Peak illness was determined to be from 5 th up to 10 th d.p.i., as seen by a large percent of total body weight lost and visibly increased breathing effort caused by severe pathological processes in the lungs ( Figure 2D).   Figure 2D) were seen. By the 6 th d.p.i., the inflammatory process had worsened only in lungs of unvaccinated mice, as characterized by interstitial pneumonia, which affected almost lungs, except the caudal sections ( Figure 2D). There were no significant differences in lung inflammation on day 6 after challenge in vaccinated mice ( Figure 2D).

Discussion
Influenza B viruses have real epidemiological significance, especially among children. Together, IBV and influenza A cause significant seasonal burdens. A lack of information about IBV's host range and the need for an adequate animal model have complicated several areas: study of pathogenicity factors and transmission methods; evaluate the antiviral drugs effect, and assessment of vaccine effectiveness. In the field of Virology, there is an entire branch devoted to obtaining recombinant strains for vaccines. Such IBV strains, however, are attenuated and apathogenic for experimental animals. Consequently, they do not provide an opportunity to study the pathological process of influenza infection, which makes it difficult to perform studies of anti-influenza drug effectiveness in vivo. Moreover, many of presented recombinant strains have lost their antigenic relevance today.
Because preclinical trials of anti-influenza drugs are conducted mainly on mice [25], we also sought a mouse model. An antigenically-relevant IBV (strain B/2017-MA) was developed, used for experimental infection in mice, and the model applied to evaluate the therapeutic and preventive effectiveness of antiviral drugs and vaccines in vivo and in vitro. Amino acid substitutions associated with IBV adaptation to mice were identified, and, probably, due to them the pathogenicity increased and enhanced replication ability in infected mammals.
On the 6 th d.p.i in BALB/c mice, influenza pneumonia featuring leukocyte and lymphocyte infiltration of bronchioles was detected in the lungs of animals infected with mouse-adapted IBV (strain B/2017-MA). On the same day, no viral loads were detected in the lung of mice infected with wild type IBV (strain B/2017). Despite this fact, pathomorphological changes were registered and characterized as mild disease.
Because the knowledge of the virions structure of influenza B remains limited [26,27] we used the electron microscopic method. We showed two shapes of virions that are not only spherical, but also elliptical. Perhaps, it is also mean that influenza B virions can have pleomorphick morphological structure.
Comparative analysis of nucleotide and amino acid substitutions between wild strain B/2017 and it's mouse-adapted variant B/2017-MA strain revealed non-synonymous nucleotide substitutions that lead to amino acid substitutions in two proteins, HA and NA. One of the amino acid substitution identified (T214I in HA) localized in the antigenically active subunit HA-HA1, thereby affecting the biological properties of the virus. Interestingly to note that same substitution was present in 0.46% of IBVs reported from 2008 to 2016 [24]. Probably, this fact may explain severe cases and poor clinical outcomes in patients infected with IBV. Another substitution was found in NA and caracterized as rare [24]. In view of the above, we assume that Asn amino acid at position 432 of the NA protein can lead to antiviral resistance and in complex with other substitution in the surface glycoprotein HA might jointly be responsible for the high pathogenicity.
Due to their roles in elevating seasonal morbidity and mortality among humans worldwide, influenza A and B viruses have epidemiological, social, and economic significance [2,13]. Selective inhibition of neuraminidase is used in the treatment of influenza to control the processes of budding and release of mature virions from the surface of an infected host cell (as a result of cleavage of sialic acid residues from hemagglutinin). In addition, NA plays a key role in the initial stages of infection, ensuring the penetration of influenza viruses into cells. Due to the specific activity of NA, inhibitors work effectively against influenza A and B viruses. Of the two NA inhibitors, oseltamivir phosphate (Tamiflu®) is considered the most effective because of its higher bioavailability (30-100%) compared to Relenza® [21]. However, the use of anti-influenza drugs to prevent and treat diseases is complicated by viral resistance which has been observed in recent years [2,29]. In addition, it has been shown that Tamiflu® a lot less effective in treating influenza B compared to influenza A [30]. Here, we show that innovative drug oseltamivir ethoxisuccinate is the promising drug in the case of strain resistance to Tamiflu. The investigational drug (oseltamivir ethoxisuccinate) is a modified version of Tamiflu, and it displays (like Tamiflu) relatively high effectiveness against the mouse-adapted IBV. Due to the fact that investigational drug (oseltamivir ethoxisuccinate) is a modified version of Tamiflu®, and its revealed high effectiveness against the mouse-adapted IBV similarly Tamiflu®, the oseltamivir ethoxisuccinate is the promising drug when dealing with resistant strains to Tamiflu®.
More attention should be paid to mammalian adaptation of IBVs and how such procedures and mechanisms can enhance the development of vaccines against current influenza strains. Due to the fact that no significant lung damage was detected in vaccinated mice at day 6 after challenge, we can conclude that Ultrix® quadrivalent vaccine is effective against our mouse-adapted IBV.

Conclusions
In this work, we serially passaged a Victoria lineage IBV 17 times in BALB/c mice. The mouse-adapted IBV caused influenza pneumonia on day 6 post inoculation. Apparently, selective accumulation of amino acid substitutions in the mouse-adapted IBV, including changes to HA (T214I) and NA (D432N), may increase pathogenicity following the adaptation and be important for the virions attach to hosts airways; however, the specific effects of these amino acid substitutions on mammalian pathogenicity requires further study. Also, were shown spherical and elliptical shapes virions of IBV.
Assessment of investigational anti-influenza drug oseltamivir ethoxisuccinate and flu vaccine Ultrix® revealed effectivity against our mouse-adapted influenza B virus. In summary, we developed adequate animal model by using antigenically-relevant mouse-adapted B/2017-MA strain for testing anti-influenza drugs and protective efficacy of flu vaccines in vitro and in vivo.