Characterization of Experimental Oro-Nasal Inoculation of Seba’s Short-Tailed Bats (Carollia perspicillata) with Bat Influenza A Virus H18N11

In 2012 and 2013, the genomic sequences of two novel influenza A virus (IAV) subtypes, designated H17N10 and H18N11, were identified via next-generation sequencing in the feces of the little yellow-shouldered fruit bat (Sturnira lilium) and the flat-faced fruit-eating bat (Artibeus planirostris), respectively. The pathogenesis caused by these viruses in their respective host species is currently insufficiently understood, which is primarily due to the inability to obtain and keep these bat species under appropriate environmental and biosafety conditions. Seba’s short-tailed bats (Carollia perspicillata), in contrast, are close relatives and a natural H18N11 reservoir species, with the advantage of established animal husbandry conditions in academic research. To study viral pathogenesis in more detail, we here oro-nasally inoculated Seba’s short-tailed bats with the bat IAV H18N11 subtype. Following inoculation, bats appeared clinically healthy, but the histologic examination of tissues revealed a mild necrotizing rhinitis. Consistently, IAV-matrix protein and H18-RNA positive cells were seen in lesioned respiratory and olfactory nasal epithelia, as well as in intestinal tissues. A RT-qPCR analysis confirmed viral replication in the conchae and intestines as well as the presence of viral RNA in the excreted feces, without horizontal transmission to naïve contact animals. Moreover, all inoculated animals seroconverted with low titers of neutralizing antibodies.


Introduction
Influenza A viruses (IAVs) originate from aquatic waterfowl and are important animal pathogens and zoonotic agents, circulating in a broad range of avian and mammalian species [1]. Although spill-over infections from animals to humans are possible, these events are exceedingly rare [2][3][4][5]. Bats (order Chiroptera) are well known as an important reservoir for various zoonotic pathogens like lyssa-, henipa-or filoviruses; however, until recently they were not considered to harbor IAVs [6,7]. This view changed in 2012 and 2013, when two novel IAVs genomes were identified by next-generation

Ethics Statement
The animal experiments described here were approved by the State Office for Agriculture, Food Safety, and Fishery of Mecklenburg-Western Pomerania under registration number LVL MV TSD/ 7221.3-1-021/18.

Virus
Animals were inoculated using recombinant H18N11 virus generated as described [18]. In brief, the eight plasmid pHW2000-based rescue system was used to generate infectious A/flat-faced bat/Peru/033/2010 (H18N11) which could be further passaged on cell culture. In detail, the recombinant bat influenza A virus H18N11 was generated by transfecting HEK293T cells (American Type Culture Collection ATCC, Manassas, VA, USA) seeded in 6-well plates with the eight pHW2000-based rescue plasmids (300 ng of each plasmid per 6-well). Forty-eight hours post transfection, the cell supernatant was collected and concentrated by ultracentrifugation through a sucrose gradient (25000 rpm, 2 h, 8 • C). Stocks were generated by infecting canine RIE1495 cells (canine epithelial cell line, stored in the "Collection of Cell Lines in Veterinary Medicine" (CCLV) at the Friedrich-Loeffler Institute in Greifswald-Insel Riems, Germany with the code CCLV-RIE 1495) at an MOI of 0.1 with concentrated HEK293T rescue supernatant for 48 h. The viral titers were determined by immunofluorescence assay, using H18-specific antibodies.

Bat Experiment and Sampling
All experiments were conducted following internal standard guidelines under biosafety level 3 conditions at the Friedrich-Loeffler-Institut. Fourteen healthy and influenza-antibody-negative (see serology section below) Seba's short-tailed bats (Carollia perspicillata) were enrolled in this study ( Figure 1A). These individuals were divided into group A (n = 4), B (n = 8) and C (n = 2). Inoculation of inhalatively anesthetized (5% isoflurane) bats was performed by the oro-nasal route with 10 5.5 tissue culture infectious dose (TCID 50 ) of H18N11 in 50 µl in two individuals of group A, and three individuals in group B. The sentinel bats of both groups were housed together with the inoculated ones 24 h after infection, in order to assess virus spread. Clinical signs (nasal discharge, reduced activity, neurological symptoms, dyspnoe), room temperature and relative humidity were monitored daily, as well as collection of pooled fecal samples. All animals in group A were sacrificed at 4 dpi and animals in group Viruses 2020, 12, 232 3 of 11 B were sacrificed at 21 dpi, followed by necropsy. Samples from the conchae, trachea, lung (left caudal lung lobe, right cranial lung lobe), heart, kidney, liver, intestine, olfactory bulb, cerebrum, cerebellum and thigh muscles were collected and stored at −80 • C until further processing for virological assays. Blood samples for serology were collected during the terminal bleeding procedure.

RNA Isolation
Organ Samples: Viral RNA extraction was achieved through solubilization of 250 µL of the supernatant of organ homogenates with 750 µl TRIzol LS Reagent (Life Technologies, Carlsbad, CA, USA). After the addition of 200 µL ROTIPURAN (Carl Roth, Karlsruhe, Germany), phase separation was attained. The following steps were completed with the NucleoMag Vet kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions on a Biosprint 96 platform (Qiagen).
Fecal samples: Viral RNA extraction of pooled fecal samples (group A pool and goup B pool) was achieved with the MasterPure™ Complete DNA and RNA Purification Kit (Lucigen, Middleton, WI, USA) according to the manufacturer's instructions, after a dilution of the samples by the factor 1:1000 in PBS.

RT-qPCR
The real-time RT-PCRs (RT-qPCR) of all organ and fecal samples were performed as described before [22]. In brief, a generic PB1 assay was used to determine the quantification cycle (Cq) using the one-step RT-qPCR Kit qScript™ XLT One-Step RT-qPCR ToughMix ® (Quantabio, Beverly, MA USA). The RT-qPCR assay was optimized for using a total volume of 12.5 µl. The reaction was run on a bio-rad cycler Cfx96 machine (Bio-Rad Laboratories, Inc. Hercules, CA, USA). Individual amplification controls on the basis of artificial spiked RNA (fecal samples, [23]) or beta actin (organ samples modified [24]) were used to evaluate inhibitory effects.

Virus Isolation
Virus isolation attempts were performed using RIE1495 cells and homogenized organ samples, which scored positive for viral RNA. Briefly, 50µL supernatant from the homogenized organ was applied onto 12.5 cm 2 cell culture flask (Corning, Corning, NY, USA). Afterwards four blind passages of potential infected cells were done, followed by a RT-qPCR based analysis.

Serology
Serum samples from all animals were heat inactivated at 56 • C for 30 min and analyzed using an indirect immunofluorescence test and a virus neutralization assay. After fixation of RIE1495 cells and RIE1495 cells infected with A/flat-faced bat/Peru/033/2010 (H18N11) using aceton methanol (1:1 vol%), the cells were incubated for one hour with the bat sera. After three washing steps using PBS, goat anti-bat IgG (H+L) secondary antibody (Novus Biologicals, Littleton, CO, USA) was applied for one hour at room temperature. After an additional three washing steps with PBS, chicken anti-goat Alexa 488 (ThermoFisher scientific, Waltham, MA, USA) was added and incubated for one hour at room temperature.
Briefly, the neutralization assay was performed with 50 µl of medium containing VSV*∆G-H18 [25] at a concentration of 10 3.3 TCID 50 that was mixed with the same volume of diluted serum. Each serum was prepared in triplicate in a 96-well plate. After incubation of 2 h at 37 • C the dilution was transferred on 100 µL medium and 24 h grown in RIE 1495 cells. The viral replication was assessed after an incubation of 5 days (37 • C, 5% CO 2 ) via visualization of GFP expression. Validation was achieved by titration of the virus dilutions.

Necropsy and Histologic Examination
A complete necropsy with macroscopic evaluation of tissues was done for all animals of this study. Histopathologic, immunohistologic and in situ hybridization workup was performed for Viruses 2020, 12, 232 5 of 11 all animals necropsied at 4 dpi (group A) and two non-inocluated bats (group C). A specimen of parenchymal organs and the skull were fixed in 4% neutral buffered formaldehyde. The skulls were decalcified (Formical 2000, Quartett Immundiagnostika und Biotechnologie Vertriebs GmbH, Berlin, Germany) and the following organs were processed to formalin fixed, paraffin embedded (FFPE) tissue blocks: four standardized coronal nasal sections with minimal adjustment to size ratios of Carollia perspicillata skulls as described elsewhere [16,18], the middle and inner ear, parotis, eye, oral cavity, esophagus, trachea, thyroidea, left caudal lung lobe, right cranial lung lobe, heart, liver, pancreas, stomach, small intestine with jejunal Peyer's patches, mesenteric lymph nodes, kidney, adrenal glands, bulbus olfactorius, cerebrum, cerebellum and bone marrow. Two to four micron-thick sections were cut and stained with hematoxylin and eosin. All specimens were evaluated for histopathologic lesions using an Axio Imager M2 microscope equipped with 10×, 20×, and 40× Plan Neofluar objectives and an AxioCam ICc3 3.3-megapixel digital camera (Carl Zeiss Microscopy GmbH, Jena, Germany).

Results
A recent study reported that H18N11 replicates especially in the lamina propria of the small intestine and the follicle-associated epithelium of the jejunal Peyer's patches of infected Jamaican fruit bats [18]. Although these acutely infected animals shed high viral loads via the rectal route, no inflammatory lesions were observed. To determine tissue tropism and pathogenesis in the related Seba's short-tailed bats (Carollia perspicillata) at different time points after infection, animals were split into three groups: Group A consisted of four bats sacrificed at 4 dpi, from which two were initially inoculated and two others that remained naïve in order to monitor viral transmission ( Figure 1A). Group B comprised three inoculated bats and five co-housed naïve contact animals that were sacrificed at 21 dpi. Two naïve bats were kept as controls in group C.
Following oro-nasal inoculation of the group A bats, none of the inoculated index or naïve contact animals exhibited clinical signs of disease and viral RNA was not present in the feces ( Figure 1B). One of two H18N11-inoculated group A individuals was found viral RNA-positive in the nasal conchae (quantification cycle value (Cq) 33.62) and in the colon (Cq 34.72) ( Table 1), whereas all other tested organ samples were PCR-negative. However, despite detection of viral RNA in some tissue samples, all attempts to isolate the infectious virus from these organs failed.
No gross lesions were observed in infected and naïve contact bats at 4 dpi, however, histopathologic examination revealed a mild, oligofocal, acute, necrotizing rhinitis in the rostral and caudal coronal sections of the nasal cavity, characterized by erosive alterations of the respiratory epithelium with pyknotic and karyorrhectic cells in the PCR-positive index bat (Figure 2A). Correspondingly, oligofocal H18-specific RNA in luminal debris ( Figure 2B) and oligofocal strong IAV-matrix protein positive cells were seen ( Figure 2C, Table 1). Erosive alterations were also present in the olfactory epithelium with mild infiltration of neutrophils in the lamina propria ( Figure 2D). Oligofocal H18-specific RNA ( Figure 2E) and oligofocal strong IAV-matrix protein positive reactions ( Figure 2F) were mainly seen in cells interpreted as sustentacular cells. In the lower respiratory tract, multifocal H18-specific RNA signals and IAV-matrix protein immunoreactive cells were seen in clusters and in single cells mostly confined to alveoli closely associated with respiratory bronchioles interpreted as alveolar macrophages and/or pneumocytes type 2 ( Figure 2I), despite a lack of unequivocal histopathologic lesions ( Figure 2G). Small intestinal villi and jejunal Peyer's patches (JPP, Figure 2J) showed no alterations. However, multifocal H18-specific RNA signals predominantly in follicle-associated epithelium (FAE) and subepithelial dome regions ( Figure 2K), and oligofocal in enteroyctes of small intestinal villi were seen in both infected bats (Table 1). A similar distribution of strong IAV-matrix protein immunoreactive cells was seen in the FAE ( Figure 2L) of JPPs and oligofocal within enterocytes and round cells in the lamina propria, interpreted as macrophages or dendritic cells ( Figure 2M). No comparable lesions, no H18-specific RNA and no IAV matrix proteins were detectable in the organs of either naïve contact animals. While the bats of group A appeared clinically healthy and shed no infectious virus until 4 dpi, two bats of group B developed a green-colored diarrhea between 9 and 21 dpi. Although we could not obtain individual rectal swabs of healthy and diseased bats without compromising their health status due to the stressful catching procedure, a RT-qPCR analysis of pooled fecal samples revealed the presence of viral RNA at 5, between 8 and 12 and 15 dpi (Figures 1B and 3). While the bats of group A appeared clinically healthy and shed no infectious virus until 4 dpi, two bats of group B developed a green-colored diarrhea between 9 and 21 dpi. Although we could not obtain individual rectal swabs of healthy and diseased bats without compromising their health status due to the stressful catching procedure, a RT-qPCR analysis of pooled fecal samples revealed the presence of viral RNA at 5, between 8 and 12 and 15 dpi ( Figure 1B and Figure 3). At necropsy, an untypical green to grey-colored and softened ingesta was seen in the colon and rectum in two of the eight bats from group B. Consistent with the observation that viral RNA was shed in the feces only until 15 dpi, the examined tissues of inoculated index and contact bats were found RT-qPCR-negative at 21 dpi. Furthermore, a histopathological analysis and immunohistochemistry of all tissues collected from animals with diarrhea (n = 2) found lesion-free, IAV-matrix protein antigen negative reactions. Importantly, aside from the green-colored softened feces, no further symptoms of disease were monitored throughout the course of the experiment.
Serum samples were taken from all group B bats after euthanasia and tested for the presence of neutralizing antibodies. The sera obtained from all three inoculated index bats revealed a low-titer seroconversion with a 50% neutralization dose (ND50) of 1:20, 1:253 and 1:16, whereas all naïve contact bats were tested seronegative ( Figure 1B).

Discussion
Here, we show that the Seba's short-tailed bat-a species showing positive seroreactivities against H18N11 in nature [9]-is susceptible to experimental oro-nasal infection with H18N11. The detection of viral RNA in tissues of the upper respiratory tract and the intestines in one of the inoculated bats at 4 dpi might indicate a productive oro-nasal and intestinal infection, although we cannot entirely exclude the possibility of detecting residual inoculum in these animals. Upon histopathological analysis, the PCR-positive bats exhibited mild necrotizing alterations confined to respiratory and olfactory epithelium. Moreover, H18-specific RNA signals and IAV-matrix protein immunoreactive cells were present in the nose and gut-associated-lymphoid-tissue (GALT) at early time points (4 dpi), and viral RNA was detected in fecal samples at later time points, which altogether At necropsy, an untypical green to grey-colored and softened ingesta was seen in the colon and rectum in two of the eight bats from group B. Consistent with the observation that viral RNA was shed in the feces only until 15 dpi, the examined tissues of inoculated index and contact bats were found RT-qPCR-negative at 21 dpi. Furthermore, a histopathological analysis and immunohistochemistry of all tissues collected from animals with diarrhea (n = 2) found lesion-free, IAV-matrix protein antigen negative reactions. Importantly, aside from the green-colored softened feces, no further symptoms of disease were monitored throughout the course of the experiment.
Serum samples were taken from all group B bats after euthanasia and tested for the presence of neutralizing antibodies. The sera obtained from all three inoculated index bats revealed a low-titer seroconversion with a 50% neutralization dose (ND50) of 1:20, 1:253 and 1:16, whereas all naïve contact bats were tested seronegative ( Figure 1B).

Discussion
Here, we show that the Seba's short-tailed bat-a species showing positive seroreactivities against H18N11 in nature [9]-is susceptible to experimental oro-nasal infection with H18N11. The detection of viral RNA in tissues of the upper respiratory tract and the intestines in one of the inoculated bats at 4 dpi might indicate a productive oro-nasal and intestinal infection, although we cannot entirely exclude the possibility of detecting residual inoculum in these animals. Upon histopathological analysis, the PCR-positive bats exhibited mild necrotizing alterations confined to respiratory and olfactory epithelium. Moreover, H18-specific RNA signals and IAV-matrix protein immunoreactive cells were present in the nose and gut-associated-lymphoid-tissue (GALT) at early time points (4 dpi), and viral RNA was detected in fecal samples at later time points, which altogether corresponds to the previously suggested gastro-intestinal tissue tropism of H18N11 in Jamaican fruit bats and its spread Viruses 2020, 12, 232 9 of 11 by the rectal route [18]. Our findings further support the idea of viral replication in MHC-II-positive antigen-presenting cells, such as macrophages and dendritic cells [16,18], which are present in high frequencies at the FAE of Peyer's patches and other inductive sites of the mucosal immunity [27].
During experimentation, some bats-presumably the index animals-developed a mild exsiccosis and catarrhal enteritis, characterized by green-colored diarrhea, and two bats additionally exhibited abnormal softened ingesta at the time of necropsy at 21 dpi. It is open to discussion whether the mild enteritis is a sequela of the IAV infection or a result of animal handling or simply an independent background lesion.
In contrast to the previous H18N11 infection study performed in Jamaican fruit bats [18], no horizontal transmission occurred among Seba's short-tailed bats. A generally unlikely, yet possible explanation for this finding could be that the contact bats were exposed to bat IAV before, but their antibody titers dropped below the detection threshold. Continuous re-exposure could have triggered a rapidly spiking and thereafter declining antibody response sufficient to prevent an infection. Such rapidly waning antibody responses after an infection-resulting in seronegativity-have been described before for Marburg virus infections in Egyptian fruit bats [28]. Another possibility could be the advanced age of the bats used in this study. Older bats mount a more efficient antiviral immune response which makes the transmission of viruses between individuals a more challenging task [29]. In this context, the previously described constitutively "always on" type I IFN activity in bat cells [30] might also interfere with viral replication and could hence restrict virus transmission among a group of bats, compared to other mammalian species. Alternatively, the lack of transmission could also be attributed to an Artibeus species specificity of H18N11. Although various South American fruit bat species have been found seropositive for H17 and H18 [9], until now full H18N11 genomic sequences or infectious viruses have been only isolated from Artibeus planirostris and Artibeus lituratus or Artibeus jamaicensis, respectively [9,18,31]. Considering this, it is tempting to speculate that species-specific differences in the immunity or the adaptation of H18 to MHC-II molecules of Artibeus spp. have influenced the infection outcome.
In summary, Seba's short-tailed bats are susceptible to, and develop, mild upper respiratory tract lesions after experimental infection with the bat-derived H18N11 subtype, but transmission to contact bats was not evident. Therefore, Seba's short-tailed bats most likely do not contribute to transmission cycles of H18N11 in nature.