Molecular basis of mammalian transmissibility of avianH1N1 influenza viruses and their pandemic potentialMark Zanina, Sook-San Wonga, Subrata Barmana, Challika Kaewborisutha,1, Peter Vogelb, Adam Rubruma,Daniel Darnella, Atanaska Marinova-Petkovaa,2, Scott Kraussa, Richard J. Webbya, and Robert G. Webstera,3

aDepartment of Infectious Diseases, St. Jude Children’s Research Hospital, Memphis, TN 38105; and bVeterinary Pathology Core, St. Jude Children’s ResearchHospital, Memphis, TN 38105

Contributed by Robert G. Webster, August 10, 2017 (sent for review March 14, 2017; reviewed by Wendy Barclay, Todd T. Davis, and Jeffrey K. Taubenberger)

North American wild birds are an important reservoir of influenza Aviruses, yet the potential of viruses in this reservoir to transmit andcause disease in mammals is not well understood. Our surveillance ofavian influenza viruses (AIVs) at Delaware Bay, USA, revealed agroup of similar H1N1 AIVs isolated in 2009, some of which wereairborne-transmissible in the ferret model without prior adaptation.Comparison of the genomes of these viruses revealed geneticmarkers of airborne transmissibility in the Polymerase Basic 2 (PB2),PB1, PB1-F2, Polymerase Acidic-X (PA-X), Nonstructural Protein 1(NS1), and Nuclear Export Protein (NEP) genes. We studied the roleof NS1 in airborne transmission and found that NS1 mutants thatwere not airborne-transmissible caused limited tissue pathology inthe upper respiratory tract (URT). Viral maturation was also delayed,evident as strong intranuclear staining and little virus at the mucosa.Our study of this naturally occurring constellation of genetic markershas provided insights into the poorly understood phenomenon ofAIV airborne transmissibility by revealing a role for NS1 and charac-teristics of viral replication in the URT that were associated withairborne transmission. The transmissibility of these viruses furtherhighlights the pandemic potential of AIVs in the wild bird reservoirand the need to maintain surveillance.

avian influenza virus | nonstructural protein 1 | H1N1 |airborne transmission | ferret

Aquatic birds are a major reservoir for influenza A viruses innature, such that 16 of the 18 hemagglutinin (HA) subtypes

and 9 of the 11 neuraminidase (NA) subtypes can be found inthis reservoir (1, 2). The North American wild bird reservoir isimportant as the causative viruses of two of the four novel humaninfluenza pandemics that have occurred in the past century haveorigins in this reservoir (3–5). Further, influenza viruses con-taining genes similar to one of these viruses, the 1918 pandemicstrain, continue to circulate in aquatic birds (6). Despite theirimportance, knowledge about the potential of the viruses in theNorth American wild bird reservoir to transmit and cause diseasein mammals is limited. Our recent studies of subtype H1N1 avianinfluenza viruses (AIVs) isolated at Delaware Bay, USA, over aperiod of more than 25 y showed that a majority of the virusesstudied caused severe disease in the mouse model (7). Moreconcerning and atypical for AIVs was that all of the viruses westudied were capable of contact transmission in the ferret modelof human influenza virus transmission (8). We also identifiedviruses that could transmit between ferrets via the airborne route(8). This was concerning as these viruses did not appear to re-quire prior adaptation to become transmissible. To investigateairborne transmission, we compared the genomes of viruses thatwere or were not airborne-transmissible in the ferret model andidentified genetic markers that correlated with transmission.These residues were in Polymerase Basic 2 (PB2), PB1, PB1-F2,Polymerase Acidic-X (PA-X), Nonstructural Protein 1 (NS1), andthe Nuclear Export Protein (NEP) (8) (Table S1). Our objectivewas to study these subtype H1N1 AIVs to better understand theairborne transmission of AIVs in mammals.

ResultsPutative Transmission Markers in Avian H1N1 Influenza Viruses Isolatedat Delaware Bay, USA, Were Important for Airborne Transmission. Webegan by confirming that the genetic markers we previously iden-tified are important for the airborne transmissibility of H1N1 AIVsbetween mammals using the ferret model (Table S1). It should benoted that we restricted these studies to wild-type viruses due tothe current pause on gain-of-function research by the US Gov-ernment. A/ruddy turnstone/Delaware/300/2009 (H1N1) (DE300),which contained all of the genetic markers and was airborne-transmissible in ferrets in our previous studies (7, 8), again trans-mitted via the airborne route. Another virus that contained all ofthe genetic markers, A/ruddy turnstone/New Jersey/AI09_256/2009(H1N1) (NJ256), also transmitted via the airborne route.A/shorebird/Delaware Bay/558/2009 (H1N1), which did not containany of these markers, did not transmit via the airborne route (Fig. 1A–C andG). Two other viruses contained some, but not all, markers.These were A/shorebird/Delaware/170/2009 (H1N1) (DE170) andA/ruddy turnstone/New Jersey/AI09_1299/2009 (H1N1) (NJ1299).Neither of these viruses showed airborne transmission, highlightingthe importance of the identified markers (Fig. 1 D, E, and G).NJ1299 lacked genetic markers in the PB2, NS1, and NEP genes,and DE170 only lacked those in the NS1 and NEP genes (Fig. 1G).

Significance

Airborne transmission of influenza A viruses between mammalsis not well understood. Our study of a group of similar subtypeH1N1 avian influenza viruses (AIVs) isolated at Delaware Bay,USA, revealed genetic markers for airborne transmission innonstructural genes. Viruses not containing all markers lackedairborne transmissibility in the ferret model. We revealed a rolefor Nonstructural Protein 1 (NS1) in facilitating airborne trans-mission that was associated with viral replication in the upperrespiratory tract (URT). Non–airborne-transmissible NS1 mutantsshowed delayed viral maturation, resulting in limited tissue pa-thology and little virus at the mucosa. Overall, our study hasrevealed insights into viral replication in the URT associated withthe airborne transmissibility of AIVs in mammals.

Author contributions: M.Z., S.-S.W., S.B., C.K., P.V., A.M.-P., R.J.W., and R.G.W. designedresearch; M.Z., S.-S.W., S.B., P.V., A.R., and A.M.-P. performed research; C.K. and S.K.contributed new reagents/analytic tools; M.Z., S.-S.W., S.B., C.K., P.V., A.R., D.D., R.J.W.,and R.G.W. analyzed data; and M.Z., R.J.W., and R.G.W. wrote the paper.

Reviewers: W.B., Imperial College; T.T.D., Centers for Disease Control and Prevention; andJ.K.T., NIAID/NIH.

The authors declare no conflict of interest.

See Commentary on page 11012.1Present address: Center for Advanced Studies for Agriculture and Food, Institute forAdvanced Studies, Kasetsart University, Bangkok 10900, Thailand.

2Present address: Influenza Division, Centers for Disease Control and Prevention, Atlanta,GA 30329.

3To whom correspondence should be addressed. Email: robert.webster@stjude.org.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1713974114/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1713974114 PNAS | October 17, 2017 | vol. 114 | no. 42 | 11217–11222

MICRO

BIOLO

GY

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

10,

202

0

Because both DE170 and NJ1299 lacked markers in NS1 and NEPand neither was airborne-transmissible, we were interested in thesegenes, particularly as they have not previously been shown to beimportant in mammalian transmissibility. As such, we then studiedthe transmissibility of a virus containing markers in the NS1 andNEP genes but lacking them in PB2, PB1, and PB1-F2: A/ruddyturnstone/New Jersey/AI_09-841/2009 (H1N1) (NJ841). This virusdid not transmit via the airborne route, suggesting that markers inNS1 and NEP are important, but not sufficient, for mammalianairborne transmissibility (Fig. 1 F and G).To determine if any mutations were present in viruses that trans-

mitted by direct contact (DC) or by airborne contact (AC), we se-quenced viruses isolated from DC and AC ferrets infected withDE300 or NJ256. We found no amino acid changes in viruses isolatedfrom infected DE300 DC or AC ferrets or NJ256 AC ferrets. In oneDC ferret infected with NJ256, we observed E204D in HA, and in theother NJ256-infected DC ferret, we observed V111L in NS1, K71N inHA, and S559L in HA. Therefore, it would appear that these virusesdid not accumulate mutations that may have contributed to airborne

transmission. There was also no evidence that differences in airbornetransmissibility were due to differences in viral load in either in-oculated or DC ferrets, as there were no significant differences in themagnitude or duration of detection of infectious virus titers in nasalwashes collected every other day beginning at 2 d postinoculation(DPI). Further, there were no obvious differences in the symptomsexhibited by ferrets inoculated with different viruses in terms of bodytemperature and weight loss (Fig. S1).

NS1 Residue 213 Was Important for Inhibition of the IFN Response.Three residues in NS1 were predicted to be important fortransmission; 7, 213, and 227. These were located in the RNAbinding domain and the C-terminal “disordered tail’ region ofthe effector domain. Residues 7 and 227 have been associatedwith viral pathogenicity in mice, but, to our knowledge, no rolehas been ascribed to residue 213 (9, 10). Segment eight of theinfluenza virus genome encodes NS1 and NEP in differentreading frames (11, 12). As the first 10 residues of NS1 and NEPare shared, NS1 7 and NEP 7 were identical. NS1 213 was within

Fig. 1. Only those viruses containing all of the putative genetic markers for airborne transmission were transmissible via the airborne route in ferrets.A/ruddy turnstone/Delaware/300/2009 (A) and A/ruddy turnstone/New Jersey/AI09_256/2009 (B) contained all genetic markers for airborne transmission andtransmitted via AC. (C) A/shorebird/Delaware/558/2009, which lacked all putative genetic markers, transmitted via DC but not via AC. A/shorebird/Delaware/170/2009 (D) and A/ruddy turnstone/New Jersey/AI09_1299/2009 (E), which contained all putative genetic markers with the exception of those in NS1 orNS1 and PB2, respectively, transmitted by DC but not via AC. (F) A/ruddy turnstone/New Jersey/AI09_841/2009 contained putative genetic markers only in PA-Xand NS1, and transmitted by DC but not via AC. Each bar represents the viral titer measured in the nasal wash from a single ferret on the indicated DPI.(G) Summary of transmission data showing putative transmissionmarkers in orange. Viruses that transmitted by AC are shaded in orange, while those that did notare shaded in green (n = 2 inoculated, DC, and AC ferrets per virus). I, inoculated.

11218 | www.pnas.org/cgi/doi/10.1073/pnas.1713974114 Zanin et al.

Dow

nloa

ded

by g

uest

on

June

10,

202

0

the overlapping coding sequences of NS1 and NEP, but in theseviruses, changes at NS1 213 did not alter NEP due to the shiftedreading frame. NS1 227 was also in the NS1/NEP overlappingcoding sequence. In these viruses, NS1 G227 corresponded toNEP G70 and NS1 E227 corresponded to NEP S70 (Table S1).NS1 has several known functions, including enhancing the

function of the influenza virus polymerase complex and interferingwith the mammalian IFN response, an important cellular antiviralresponse (13). As such, we studied these activities of DE300 andDE170 NS1. We performed polymerase assays using the poly-merase complex of DE300 and the NS1 of DE300 or DE170 todetermine if there was a difference between the NS1 genes ofairborne-transmissible and non–airborne-transmissible viruses.There were no significant differences between the polymeraseactivity of this complex coexpressed with DE170 or DE300 NS1(Fig. S2). Interestingly, the NS1 protein of DE300 was less ableto inhibit the retinoic acid-inducible gene 1 (RIG-I)–mediatedactivation of the IFN-β promoter in vitro compared with theDE170 NS1 (Fig. 2). We mutated the residues at these positionsin DE300 NS1 to those in DE170 (i.e., L7S, S213P, G227E). L7Sand G227E did not have any effect on the activity of DE300 NS1;however, S213P alone reverted the activity of DE300 to that ofDE170 (Fig. 2). Therefore, residue 213 appeared critical for thisfunction of NS1.

Mutations in NS1 Abrogated Airborne Transmissibility and AlteredViral Replication and Pathology in the Upper Respiratory Tract. Ourtransmission studies revealed that markers in NS1/NEP wererequired, but not sufficient, for airborne transmission (Fig. 1). Aswe were restricted from conducting gain-of-function research, weused a loss-of-function strategy to study the role of NS1/NEP inairborne transmission using viruses produced via reverse genetics(rg) (14). We constructed an rg version of DE300 (rgDE300),which was airborne-transmissible (Fig. 1A). As position 213 inNS1 appeared important for the suppression of the IFN pathway(Fig. 2), we introduced NS1 S213P into rgDE300 as a loss-of-

function mutation to produce the rgDE300 P213 virus. We alsostudied the transmissibility of A/gull/Delaware/AI09_438/2009(H1N1) (DE438), which contained all of the markers in PB2,PB1, PB1-F2, and PA-X. In NS1/NEP, DE438 only containedthe marker S213 in NS1 (Fig. 3D). Therefore, we assessed theimportance of position 213 in NS1 for airborne transmissionusing a virus containing all transmission markers except NS1 213(rgDE300 P213) and a virus containing no transmission markersin NS1/NEP with the exception of NS1 S213 (DE438).There were no significant differences in the infectious virus

titers in nasal washes obtained from ferrets inoculated withrgDE300, rgDE300 P213, or DE438 and rgDE300 transmittedvia the airborne route in a similar fashion to DE300 (Fig. 3A).However, both rgDE300 P213 and DE438 were not airborne-transmissible (Fig. 3 B and C). DE438 was similar to rgDE300P213 in that there was a lack of transmission to AC ferrets, andthe nasal washes of only one DC ferret contained infectious virustiters until 6 DPI (Fig. 3C). Again there were no obvious dif-ferences in the symptoms exhibited by these ferrets in terms ofbody temperature or weight loss (Fig. S3). Further, these virusesshowed similar in vitro growth kinetics (Fig. S4). Therefore, NS1S213 was clearly important for airborne transmission; however,in the context of NS1/NEP, NS1 S213 alone was insufficient forairborne transmissibility.We next studied the respiratory tract of ferrets inoculated with

rgDE300, rgDE300 P213, or DE438 to determine if there wereany correlations between viral spread and/or pathology thatcorrelated with airborne transmission. We studied these tissuesat 5 DPI, 1 d after infectious virus titers were detected in thenasal washes of rgDE300 ACs (Fig. 3A). In the lower respiratorytract, we observed clusters of infected bronchiolar epithelial cellsin many lobes of the lung in all inoculated ferrets, with limitedspread to alveoli. There were no significant differences betweenferrets inoculated with different viruses. However, influenza vi-rus staining of the nasal respiratory epithelium and olfactoryneuroepithelium appeared to be more persistent and extensivefollowing inoculation with rgDE300 P213 or DE438, neither ofwhich was airborne-transmissible, compared with rgDE300 (Fig.4 A–C). Moreover, rgDE300 P213 and DE438 were found pre-dominantly in the nucleus of infected cells in contrast torgDE300, which was more evenly distributed in both the nucleusand cytoplasm. Immunostaining also revealed less virus anti-gen at the mucosal surface in ferrets inoculated with rgDE300P213 or DE438 compared with ferrets inoculated with rgDE300(Fig. 4 A–C). The rgDE300 P213- and DE438-infected cellsappeared rounded up and exfoliated; however, apoptosis/ne-crosis was rare in comparison to infected cells in the rgDE300-infected mucosa, in which they were numerous (Fig. 4 D–I).NS1 has been shown to influence apoptosis in vitro, althoughboth apoptotic and antiapoptotic roles have been observed (15,16). Overall, these data suggest that differences in intracellularmaturation in infected mucosa may have resulted in less virus atthe nasal mucosal surfaces in ferrets inoculated with rgDE300P213 or DE438, compared with ferrets inoculated with rgDE300.However, infectious virus titers in nasal washes from ferrets in-oculated with these viruses were similar. This is suggestive of adisconnect between airborne transmissibility and infectious viraltiters in nasal washes.As NS1 is known to modulate the immune response, we exam-

ined the expression of cytokines and chemokines in the respiratorytract of inoculated ferrets using quantitative RT-PCR. Thereappeared to be greater expression of IFN-α and IFN-β mRNA inthe nasal turbinates of ferrets inoculated with rgDE300 comparedwith ferrets inoculated with rgDE300 P213 or DE438 (Fig. S5 Aand B). The expression of IFN-γ mRNA also appeared greater inthe nasal turbinates of rgDE300- and rgDE300 P213-inoculatedferrets compared with DE438-inoculated ferrets (Fig. S5C). Theexpression of these IFNs in the tracheas of these animals appeared

Fig. 2. Residue 213 in the NS1 of DE300 was critical for the inhibition of theIFN pathway. The NS1 of DE170, which was not airborne-transmissible inferrets, inhibited the IFN-β promoter significantly more than the NS1 ofDE300, which was airborne-transmissible in ferrets. Residue 213 was criticalfor this difference, as L7S and G227E did not alter the activity of DE300NS1 but S213P changed the activity of DE300 NS1 to that of DE170 NS1. Themean ± SEM is shown. Values are normalized to the activity of the IFN-βpromoter in the absence of NS1 (i.e., IFN-β promoter activity without in-hibition). ***P < 0.001.

Zanin et al. PNAS | October 17, 2017 | vol. 114 | no. 42 | 11219

MICRO

BIOLO

GY

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

10,

202

0

similar (Fig. S5 D–F). Further, there were also no significant dif-ferences in the expression of selected cytokines/chemokines in fourdifferent lobes of the lung (Fig. S6). While these data are sugges-tive of a correlation between IFN expression in the upper re-spiratory tract (URT) and airborne transmission, variability andsmall sample sizes (n = 2) limited analysis of these data.

H1N1 AIVs Containing All Putative Genetic Markers for TransmissionWere Only Isolated at Delaware Bay, USA, in 2009. We isolated 36H1N1 AIVs from shorebirds during our surveillance at DelawareBay, USA, during the period of 1990–2015, but the majority ofthese viruses were isolated in 2009 (17 H1N1 AIVs). Further,putative transmission markers were only common in H1N1 AIVsisolated in 2009 (Table S1). Only one marker, G31 in PB1-F2,was present in H1N1 AIVs isolated after 2009. To investigate theorigins of these markers, we conducted phylogenetic analysis onthese 36 H1N1 AIVs to determine if gene segments in airborne-transmissible viruses had different origins compared with thosein non–airborne-transmissible viruses. Overall, the gene seg-ments found in airborne-transmissible AIVs did not cluster dis-tinctly from those found in non–airborne-transmissible AIVs(Fig. S7). Further, there were no indications that gene segmentsassociated with airborne transmission were phylogenetically re-lated to those of prior pandemic H1N1 viruses (Fig. S7).

DiscussionAirborne transmission between humans is an important phenotypeof pandemic influenza viruses that is incompletely understood. Ourstudy has shown that AIVs in the wild bird reservoir are capable ofairborne transmission in mammals without prior adaptation in anintermediate host, indicating that the pandemic potential of theseviruses is greater than currently appreciated. We have also studieda role for NS1 in facilitating mammalian airborne transmission thatis linked to aspects of viral replication in the URT that, to ourknowledge, have not been previously characterized.Several physiological differences, or species barriers, in avian

species that limit AIV infection and transmission in humans havebeen identified. These include the higher temperature and the

expression of different sialic acid linkages in the avian en-teric tract, the primary site of viral replication in avian species,compared with the human respiratory tract (17–19). Well-characterized adaptations in the influenza virus surface glyco-protein genes HA and NA and in the internal PB2 gene havebeen linked with overcoming these species barriers (20–24).These include E627K in PB2, associated with replication in themammalian respiratory tract, and E190D and G225D in HA(H3 numbering), associated with preferential binding to human-type sialic acid linkages (25–28). The viruses studied here did notcontain these mutations. Further, the transmission markersidentified in these viruses were not in the surface glycoproteinsof the virus and did not include other residues previously linkedto mammalian transmissibility of AIVs. Further, mutationsknown to be associated with changes in HA sialic acid bindingpreferences were not detected in the consensus sequence of theHA gene of DE300 viruses isolated from infected AC ferrets. Wepreviously demonstrated that DE300 showed binding to bothα2,3- and α2,6-linked sialic acids and that selection for α2,6-linked sialic acid binding seemed to occur with airborne trans-mission (8). Surprisingly, we did not detect any mutations in HAindicative of changes in sialic acid binding preferences usingSanger sequencing. Therefore, it is possible that changes in theminor variant populations may have impacted the overall sialicacid binding preferences of DE300 isolated from AC ferrets.Future studies using deep sequencing are likely to reveal if this isthe case. Overall, the determinants of airborne transmissibility ofthese viruses appeared to be outside those classically associatedwith this phenotype.NS1 was a critical component of the airborne transmission

phenotype of these viruses, such that a single residue change wassufficient to abrogate airborne transmission. Owing to their im-munomodulatory activities, NS1, PB1-F2, and PA-X have beenlinked to improving replication in mammals, but not directlylinked to airborne transmission (29–32). Here, we have shownthat NS1 had a clear impact on viral replication and pathology inthe URT that was likely important for airborne transmission.Overall, there appeared to be some impediment to the intracellular

Fig. 3. NS1 was critical for AC, but not DC, transmission in ferrets. (A) The rg virus A/ruddy turnstone/Delaware/300/2009 (H1N1) (rgDE300) transmitted to DCand AC ferrets, as per the wild-type DE300 virus. (B) Introduction of S213P into the NS1 of rgDE300 abrogated AC, but not DC, transmission. (C) NS1 of A/gull/Delaware/AI09_438/2009 (H1N1), which contained S213 in conjunction with S7 and E227, transmitted to DC ferrets, but not to AC ferrets. Each bar representsthe viral titer measured in the nasal wash from a single ferret on the indicated DPI. (D) Summary of transmission data showing putative transmission markersin orange. Viruses that transmitted by AC are shaded in orange, while those that did not are shaded in green (n = 2 inoculated, DC, and AC ferrets per virus). I,inoculated.

11220 | www.pnas.org/cgi/doi/10.1073/pnas.1713974114 Zanin et al.

Dow

nloa

ded

by g

uest

on

June

10,

202

0

maturation of non–airborne-transmissible viruses that resulted inless virus at the mucosal surface in the URT. However, these dif-ferences were not reflected in the magnitude or duration of de-tection of infectious virus titers in the nasal washes of inoculatedferrets, suggesting that another aspect of viral shedding may havebeen affected.In summary, we have demonstrated the importance of these

markers for airborne transmission. We have also identified a rolefor NS1 as a facilitator of airborne transmission in these viruses andcharacterized how mutations in NS1 impacted viral replication inthe URT. That this naturally occurring constellation of gene seg-ments conferred airborne transmissibility to these viruses withoutadaptation further reinforces the pandemic potential of AIVs inthe wild bird reservoir and the need to maintain surveillance.

Experimental ProceduresViruses and Cells. Viruses used to inoculate ferrets were minimally passaged in10-d-old embryonated chicken eggs (7). Virus titers were determined by

measuring the egg infectious dose, 50% (EID50). The viruses generated by rgin this study were produced using the eight-plasmid rg system and propa-gated in eggs as described previously (14). Viruses generated by rg weresequenced to ensure that there were no sequence differences comparedwith the wild-type virus. The A549 cells were maintained in Kaighn’s Mod-ification of Ham’s F-12 Medium (American Type Culture Collection) con-taining 10% FBS (Invitrogen), L-glutamine (Invitrogen), and penicillin/streptomycin (Invitrogen). The 293T cells were maintained in OptiMEMcontaining 5% FBS (Invitrogen). Sanger sequencing was used to determine ifany mutations were present in viruses isolated from ferrets compared withthe virus stock used for inoculation.

Ferret Experiments. Three- to four-month-old outbred male ferrets (Mustelaputorius furo) were purchased from Triple F Farms. All experiments wereconducted in an animal biosafety level 2+ (ABSL2+) facility in compliancewith the policies of the National Institutes of Health and the Animal WelfareAct and with the approval of the St. Jude Children’s Research Hospital In-stitutional Animal Care and Use Committee (protocol no. 428, approvedSeptember 2015). Inoculated ferrets were anesthetized using 3% isoflurane(2% oxygen) and inoculated intranasally with 106 EID50 of virus in 1 mL of

Fig. 4. Virus at the mucosal surfaces of the respiratory epithelium and olfactory neuroepithelium was critical for airborne transmission. More virus stainingwas observed at the mucosal surfaces of the nasal respiratory epithelium and olfactory neuroepithelium of ferrets 5 DPI with the rg virus A/ruddy turnstone/Delaware/300/2009 (H1N1) (rgDE300) (A) compared with ferrets inoculated with rgDE300 containing P213 in NS1 (rgDE300 P213) (B) or A/gull/Delaware/AI09_438/2009 (H1N1) (DE438) (C). The rgDE300 P213 and DE438 were found predominantly in the nucleus of infected cells, while rgDE300 was found in thecytoplasm and the nucleus. Apoptotic/necrotic cells were numerous in these tissues in ferrets inoculated with rgDE300 (D and G), while they were rare inferrets inoculated with rgDE300 P213 (E and H) or DE438 (F and I). Sections were stained with anti-H1N1 influenza virus antibody (brown staining) (A–C), withhemotoxin and eosin (D–F), or with caspase 3 (brown staining) and hematoxylin (blue staining) (G–I) (n = 2 ferrets per virus). (Scale bars: 50 μm.)

Zanin et al. PNAS | October 17, 2017 | vol. 114 | no. 42 | 11221

MICRO

BIOLO

GY

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

10,

202

0

PBS. At 1 DPI, one DC ferret was placed in the same cage with each in-oculated ferret and one AC ferret was placed in an adjacent cage separatedby a wire grill. Six ferrets were used per virus: two inoculated, two DCs, andtwo ACs. All ferrets were monitored daily for weight loss, body temperature,and clinical signs of influenza for 16 DPI. Every second day, ferrets werelightly anesthetized with 40 mg/kg ketamine administered intramuscularly,and nasal wash specimens were collected using 1 mL of sterile PBS as de-scribed previously (7). Viral titers in nasal wash specimens were determinedby EID50.

Histology and Immunohistochemistry. Ferrets were euthanized according toinstitutional protocols. Whole lungs, tracheas, and nasal turbinates werefixed in 10%neutral buffered formalin, embedded in paraffin, and sectioned.Sections were stained using hematoxylin and eosin for histology or anantiinfluenza virus H1N1 polyclonal antibody (US Biologicals) for immuno-histochemistry. In vitro immunohistochemistry performed on Madin–Darbycanine kidney cells infected with these viruses did not reveal any differencesin the specificity of this antibody. Cleaved caspase 3 was detected followingantigen retrieval at 100 °C for 32 min on a Discovery Ultra immunostainer(Ventana Medical Systems). Slides were stained with a rabbit polyclonalanti–caspase-3 antibody (Biocare Medical), followed by goat anti-rabbitpolyclonal antibodies labeled with a proprietary hapten nitropyrazole(OmniMap anti-Rb HRP; Ventana Medical Systems). Color was then de-veloped with ChromoMap 3,3′-diaminobenzidine tetrahydrochloride chro-mogen, and slides were then counterstained with hematoxylin (all fromVentana Medical Systems). Sections were scored in a blinded fashion.

IFN-β Promoter Activity Assay. Plasmids expressing the IFN-β promoter, fireflyluciferase, the Renilla luciferase plasmid reporter containing a thymidine kinase

promoter (Promega), RIG-I, and the NS1 gene in question were cotransfected in293T cells using XtremeGene HP (Promega) (33). Cells were harvested at 24 hposttransfection, and luciferase activity was measured using the Dual-LuciferaseReporter Assay kit (Promega) as per the manufacturer’s instructions. Fireflyluciferase activity was normalized to that of Renilla luciferase. Experimentswere performed in triplicate.

Quantitative PCR. RNA was extracted from ferret tissues or cells in TRIzol(Invitrogen) using the Direct-zol RNA MiniPrep Kit (Zymo Research). One mi-crogram of RNA was used for RT using SuperScript III reverse transcriptase(Invitrogen) and random primers (Invitrogen). Quantitative PCR was performedusing SYBR Green PCR master mix (QIAGEN) and a 7500 Fast Quantitative PCRSystem (Applied Biosystems). The housekeeping gene β-actin was used as anendogenous control, and relative quantification values were calculated bycomparing inoculated groups with uninoculated groups.

Statistical Analysis. Data were graphed and analyzed using GraphPad Prism(GraphPad Software). Statistical analyses were performed by ANOVA. Pvalues of <0.05 were deemed statistically significant.

ACKNOWLEDGMENTS. We thank Professor David Stallknecht and RebeccaPoulson for generously sharing with us some of the viruses studied here. Wethank Jeri-Carol Crumpton, Jennifer DeBeauchamp, and John Franks forexpert technical assistance. We thank the staff of the Animal ResourcesCenter at St. Jude Children’s Research Hospital for taking excellent care ofthe animals. We thank the Veterinary Pathology Core at St. Jude Children’sResearch Hospital for their work on the pathology studies. This study wasfunded by the National Institute of Allergy and Infectious Diseases and theNIH under Contracts HHSN266200700005C and HHSN272201400006C.

1. Olsen B, et al. (2006) Global patterns of influenza a virus in wild birds. Science 312:

384–388.2. Webster RG, Bean WJ, Gorman OT, Chambers TM, Kawaoka Y (1992) Evolution and

ecology of influenza A viruses. Microbiol Rev 56:152–179.3. Scholtissek C, von Hoyningen V, Rott R (1978) Genetic relatedness between the new

1977 epidemic strains (H1N1) of influenza and human influenza strains isolated be-

tween 1947 and 1957 (H1N1). Virology 89:613–617.4. Smith GJ, et al. (2009) Origins and evolutionary genomics of the 2009 swine-origin

H1N1 influenza A epidemic. Nature 459:1122–1125.5. Taubenberger JK, Reid AH, Krafft AE, Bijwaard KE, Fanning TG (1997) Initial genetic

characterization of the 1918 “Spanish” influenza virus. Science 275:1793–1796.6. Watanabe T, et al. (2014) Circulating avian influenza viruses closely related to the

1918 virus have pandemic potential. Cell Host Microbe 15:692–705.7. Koçer ZA, Krauss S, Stallknecht DE, Rehg JE, Webster RG (2012) The potential of avian

H1N1 influenza A viruses to replicate and cause disease in mammalian models. PLoS

One 7:e41609.8. Koçer ZA, et al. (2015) Possible basis for the emergence of H1N1 viruses with pan-

demic potential from avian hosts. Emerg Microbes Infect 4:e40.9. Koçer ZA, et al. (2014) Survival analysis of infected mice reveals pathogenic variations

in the genome of avian H1N1 viruses. Sci Rep 4:7455.10. Shin YK, et al. (2007) SH3 binding motif 1 in influenza A virus NS1 protein is essential

for PI3K/Akt signaling pathway activation. J Virol 81:12730–12739.11. Inglis SC, Barrett T, Brown CM, Almond JW (1979) The smallest genome RNA segment

of influenza virus contains two genes that may overlap. Proc Natl Acad Sci USA 76:

3790–3794.12. Lamb RA, Choppin PW (1979) Segment 8 of the influenza virus genome is unique in

coding for two polypeptides. Proc Natl Acad Sci USA 76:4908–4912.13. García-Sastre A, et al. (1998) Influenza A virus lacking the NS1 gene replicates in

interferon-deficient systems. Virology 252:324–330.14. Hoffmann E, Neumann G, Kawaoka Y, Hobom G, Webster RG (2000) A DNA trans-

fection system for generation of influenza A virus from eight plasmids. Proc Natl Acad

Sci USA 97:6108–6113.15. Schultz-Cherry S, Dybdahl-Sissoko N, Neumann G, Kawaoka Y, Hinshaw VS (2001)

Influenza virus ns1 protein induces apoptosis in cultured cells. J Virol 75:7875–7881.16. Zhirnov OP, Klenk HD (2007) Control of apoptosis in influenza virus-infected cells by

up-regulation of Akt and p53 signaling. Apoptosis 12:1419–1432.17. Scull MA, et al. (2009) Avian influenza virus glycoproteins restrict virus replication and

spread through human airway epithelium at temperatures of the proximal airways.

PLoS Pathog 5:e1000424.

18. Nicholls JM, Bourne AJ, Chen H, Guan Y, Peiris JS (2007) Sialic acid receptor detectionin the human respiratory tract: evidence for widespread distribution of potentialbinding sites for human and avian influenza viruses. Respir Res 8:73.

19. Shinya K, et al. (2006) Avian flu: Influenza virus receptors in the human airway.Nature 440:435–436.

20. Herfst S, et al. (2012) Airborne transmission of influenza A/H5N1 virus between fer-rets. Science 336:1534–1541.

21. Imai M, et al. (2012) Experimental adaptation of an influenza H5 HA confers re-spiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets. Nature486:420–428.

22. de Wit E, et al. (2010) Molecular determinants of adaptation of highly pathogenicavian influenza H7N7 viruses to efficient replication in the human host. J Virol 84:1597–1606.

23. Liu Y, et al. (2010) Altered receptor specificity and cell tropism of D222G hemag-glutinin mutants isolated from fatal cases of pandemic A(H1N1) 2009 influenza virus.J Virol 84:12069–12074.

24. Tumpey TM, et al. (2007) A two-amino acid change in the hemagglutinin of the1918 influenza virus abolishes transmission. Science 315:655–659.

25. Matrosovich M, et al. (2000) Early alterations of the receptor-binding properties ofH1, H2, and H3 avian influenza virus hemagglutinins after their introduction intomammals. J Virol 74:8502–8512.

26. Hatta M, Gao P, Halfmann P, Kawaoka Y (2001) Molecular basis for high virulence ofHong Kong H5N1 influenza A viruses. Science 293:1840–1842.

27. Hatta M, et al. (2007) Growth of H5N1 influenza A viruses in the upper respiratorytracts of mice. PLoS Pathog 3:1374–1379.

28. Subbarao EK, London W, Murphy BR (1993) A single amino acid in the PB2 gene ofinfluenza A virus is a determinant of host range. J Virol 67:1761–1764.

29. Conenello GM, Zamarin D, Perrone LA, Tumpey T, Palese P (2007) A single mutation inthe PB1-F2 of H5N1 (HK/97) and 1918 influenza A viruses contributes to increasedvirulence. PLoS Pathog 3:1414–1421.

30. Jiao P, et al. (2008) A single-amino-acid substitution in the NS1 protein changes thepathogenicity of H5N1 avian influenza viruses in mice. J Virol 82:1146–1154.

31. Seo SH, Hoffmann E, Webster RG (2002) Lethal H5N1 influenza viruses escape hostanti-viral cytokine responses. Nat Med 8:950–954.

32. Zhou H, et al. (2010) Effect on virulence and pathogenicity of H5N1 influenza A virusthrough truncations of NS1 eIF4GI binding domain. J Infect Dis 202:1338–1346.

33. Tawaratsumida K, et al. (2014) Quantitative proteomic analysis of the influenza Avirus nonstructural proteins NS1 and NS2 during natural cell infection identifies PACTas an NS1 target protein and antiviral host factor. J Virol 88:9038–9048.

34. Reed LJ, Muench H (1938) A simple method of estimating fifty percent endpoints. AmJ Hyg 27:493–497.

11222 | www.pnas.org/cgi/doi/10.1073/pnas.1713974114 Zanin et al.

Dow

nloa

ded

by g

uest

on

June

10,

202

0