Prion protein protects mice from lethal infection with influenza A viruses

The cellular prion protein, designated PrPC, is a membrane glycoprotein expressed abundantly in brains and to a lesser extent in other tissues. Conformational conversion of PrPC into the amyloidogenic isoform is a key pathogenic event in prion diseases. However, the physiological functions of PrPC remain largely unknown, particularly in non-neuronal tissues. Here, we show that PrPC is expressed in lung epithelial cells, including alveolar type 1 and 2 cells and bronchiolar Clara cells. Compared with wild-type (WT) mice, PrPC-null mice (Prnp0/0) were highly susceptible to influenza A viruses (IAVs), with higher mortality. Infected Prnp0/0 lungs were severely injured, with higher inflammation and higher apoptosis of epithelial cells, and contained higher reactive oxygen species (ROS) than control WT lungs. Treatment with a ROS scavenger or an inhibitor of xanthine oxidase (XO), a major ROS-generating enzyme in IAV-infected lungs, rescued Prnp0/0 mice from the lethal infection with IAV. Moreover, Prnp0/0 mice transgenic for PrP with a deletion of the Cu-binding octapeptide repeat (OR) region, Tg(PrPΔOR)/Prnp0/0 mice, were also highly susceptible to IAV infection. These results indicate that PrPC has a protective role against lethal infection with IAVs through the Cu-binding OR region by reducing ROS in infected lungs. Cu content and the activity of anti-oxidant enzyme Cu/Zn-dependent superoxide dismutase, SOD1, were lower in Prnp0/0 and Tg(PrPΔOR)/Prnp0/0 lungs than in WT lungs. It is thus conceivable that PrPC functions to maintain Cu content and regulate SOD1 through the OR region in lungs, thereby reducing ROS in IAV-infected lungs and eventually protecting them from lethal infection with IAVs. Our current results highlight the role of PrPC in protection against IAV infection, and suggest that PrPC might be a novel target molecule for anti-influenza therapeutics.


Introduction
The normal cellular prion protein, designated PrP C , is a membrane glycoprotein tethered to the outer cell membrane via a glycosylphosphatidylinositol anchor moiety and expressed most abundantly in brains, particularly by neurons, and to a lesser extent in non-neuronal tissues including hearts, kidneys, and lungs [1,2]. Conformational conversion of PrP C into the abnormally folded, amyloidogenic isoform is a pivotal pathogenic event in prion diseases, a group of neurodegenerative disorders including Creutzfeldt-Jakob disease in humans and scrapie and bovine spongiform encephalopathy in animals [2]. In brains, glial cells including microglia, astrocytes, and oligodendrocytes also express PrP C [3][4][5][6][7]. PrP C expression has also been reported in non-cardiomyocytes in hearts [8], in glomeruli, proximal convoluted tubules and collecting ducts in kidneys [8,9], activated hepatic stellate cells in livers [10], in lymphoid nodules [11] and perilymphoid zones of the red pulp [8] in spleens, and in neuronal cells in the lamina propia and parasympathetic ganglions [8], some epithelial cells [12], Peyer's patches [12] and enteric glial cells [13] in intestines. In lungs, alveolar walls were reported to be positive for PrP C expression [8]. However, the exact function of PrP C remains to be clarified.
Neuroprotective function has been suggested for PrP C . Mice devoid of PrP C (Prnp 0/0 ) have been reported to be vulnerable to ischemic brain injury, with enhanced neuronal cell apoptosis in the injured brains [14][15][16]. PrP lacking the octapeptide repeat (OR) region failed to rescue Prnp 0/0 mice from ischemic brain injury [17]. These results suggest that PrP C might exert an anti-apoptotic activity through the OR region, thereby protecting neurons from ischemic damage. It was recently reported that the hearts and kidneys of Prnp 0/0 mice were also vulnerable to ischemic injury [18,19], indicating that PrP C could have a protective function even in nonneuronal tissues. However, the exact mechanism underlying the protective function of PrP C remains elusive.
The OR region binds to Cu ions via histidine residues [20][21][22]. Some investigators showed that Cu content was reduced and the enzymatic activity of anti-oxidant enzyme Cu/Zn-dependent superoxide dismutase, SOD1, was lower in the brains of Prnp 0/0 mice [20,23,24], suggesting that PrP C might function to maintain Cu levels, thereby regulating SOD1 activity to exert anti-oxidative activity and eventually protecting neurons from apoptosis. However, others reported normal levels of Cu content and SOD1 activity in the brains of Prnp 0/0 mice [25]. Thus, the role of PrP C in maintenance of Cu content and regulation of SOD1 in terms of its protective activity remains to be determined.
Several groups have investigated the role of PrP C in virus infection in mice [26][27][28]. Nasu-Nishimura et al. reported that PrP C could have a protective role against infection with encephalomyocarditis virus B variant by reducing neuronal apoptosis in the brains of infected mice without affecting viral replication [27]. It was also reported that human immunodeficiency virus type 1 (HIV-1) production was strongly inhibited by expression of PrP C in cultured cells transfected with an infectious HIV-1 molecular clone [28]. On the other hand, PrP C overexpression was shown to enhance acute infection of herpes simplex virus type 1 (SC16) in the central and peripheral neuronal tissues, causing higher mortality in mice, although latent infection of the virus in these tissues was inhibited by overexpression of PrP C [26].
Influenza A virus (IAV) is an enveloped, negative sense, single-stranded RNA virus, causing seasonal epidemic outbreaks of influenza [29]. High morbidity and mortality are observed in infected people, particularly in the young and elderly and those with underlying chronic diseases in lung or cardiovascular systems [29]. Several lines of evidence indicate that reactive oxygen species (ROS) play a pivotal role in IAV infection-induced lung injury, by causing apoptosis in infected lung epithelial cells [30][31][32][33]. However, the role of PrP C in IAV infection remains unknown.
In the present study, we show that Prnp 0/0 mice were highly susceptible to IAV infection, with higher mortality, compared to wild-type (WT) mice. PrP lacking the Cu-binding OR region failed to rescue Prnp 0/0 mice from lethal infection with IAV. Infected Prnp 0/0 lungs were severely injured, with higher epithelial cell apoptosis and higher ROS levels than control WT lungs. Treatment with anti-oxidants rescued Prnp 0/0 mice from lethal infection with IAV. SOD1 activity and Cu ion content were lower in Prnp 0/0 lungs than in WT lungs. These results suggest that PrP C could have a protective role against lethal infection with IAVs through the OR region probably by exerting an anti-oxidative activity by maintaining Cu content and regulating SOD1 in lungs.

PrP C is expressed in lung epithelial cells
We first investigated expression of PrP C in lung tissues of C57BL/6 WT mice on Western blotting. PrP C was detectable in various tissues, with highest expression in brains ( Fig 1A). Lower but considerably high levels of PrP C were detected in lungs, followed by that in spleens and intestines ( Fig 1A). Only very low level of PrP C was detectable in hearts and livers ( Fig 1A). These results are consistent with PrP C being expressed most abundantly in brains and, to lesser extents, in other non-neuronal tissues [1,2]. Weak signals were observed in Prnp 0/0 lungs ( Fig  1A). However, no signals for PrP C were detectable in the brains of Prnp 0/0 mice (Fig 1A), clearly indicating that PrP C expression is absent in Prnp 0/0 mice. Therefore, the signals observed in Prnp 0/0 lungs are not specific for PrP C . PrP C is a glycoprotein with two glycosylation sites, therefore di-, mono-, and un-glycosylated forms of PrP C are being expressed and detected as a broad band on Western blotting. We then performed immunofluorescent staining for PrP C in lungs. No specific signals were detected on Prnp 0/0 lung slices (Fig 1B). In contrast, bronchiolar and alveolar epithelial cells on WT lung slices showed positive staining ( Fig  1B). Double staining with anti-podoplanin, anti-surfactant protein C (SP-C), or anti-Clara cell 10-kDa protein (CC10) antibodies, which specifically detect alveolar type 1 and 2 epithelial cells (AT1 and AT2 cells) and bronchiolar Clara epithelial cells, respectively, revealed expression of PrP C in these lung epithelial cells (Fig 1C).
higher sensitivity to IAV/PR8, with markedly elevated mortality (Fig 2A). At 14 days postinfection (dpi), only about 7% of male Prnp 0/0 mice survived the infection while more than 70% of male WT mice were still alive. Higher mortality was also observed in infected female Prnp 0/0 mice, compared to control female WT mice (Fig 2A). Viral titers were higher in infected Prnp 0/0 lungs than in control WT lungs (Fig 2A). Western blotting showed similar expression of PrP C in male and female WT lungs (S1 Fig). We also intranasally infected male Prnp 0/0 and WT mice with increasing infectious doses (100 IFU) of IAV/PR8. None of Prnp 0/0 mice survived the infection by 14 dpi (Fig 2B). However, about 40% of WT mice remained alive ( Fig 2B). Viral titers were higher in infected Prnp 0/0 lungs than in control WT lungs ( Fig  2B). According to the Reed and Muench method [34], a 50% mouse lethal dose (MLD50) for IAV/PR8 was calculated as 66 IFU in WT mice and less than 50 IFU in Prnp 0/0 mice. We also Western blotting of various tissues from C57BL/6 WT and Prnp 0/0 mice with 6D11 anti-PrP antibody, which recognizes residues 93-109 of mouse PrP. Non-specific weak signals were observed in lungs and spleens. Actb is an internal control. Br, brain; Ht, heart; Lg, lung; Lv, liver; Sp, spleen; In, intestine. (B) Immunofluorescence staining of WT and Prnp 0/0 lungs with IBL-N anti-PrP antibodies, which are raised against a synthetic N-terminal peptide. Bar, 400 μm. Insets show 2 times-magnified images of white squares. (C) Double immunofluorescence staining of WT and Prnp 0/0 lungs with IBL-N anti-PrP antibodies and antibodies against podoplanin, SP-C, or CC10. Bar, 400 μm. Insets show 2.5 times-magnified images of white squares.  used Prnp 0/0 and Prnp +/+ littermates for intranasal infection with 50 IFU of IAV/PR8. Male and female Prnp +/+ mice showed a mortality rate of about 40% at 14 dpi ( Fig 2C). However, more than 80% of male and female Prnp 0/0 mice died by 14 dpi (Fig 2C). Higher viral titers were observed in infected Prnp 0/0 lungs compared to control Prnp +/+ lungs (Fig 2C).
We also tested other IAV strains, A/Aichi/2/68 (H3N2) and A/WSN/33 (H1N1) (hereafter referred to as IAV/Aichi and IAV/WSN, respectively), for their pathogenicity in Prnp 0/0 mice. Prnp 0/0 and WT mice were intranasally infected with 500 IFU of IAV/Aichi and 3,000 IFU of IAV/WSN. IAV/WSN belong to the same H1N1 subtype family as IAV/PR8. However, IAV/ WSN was established by passages in mouse brains, thus being neurotropic, while IAV/PR8 is highly pathogenic to lungs [35]. Therefore, higher virus titers were used for intranasal infection with IAV/WSN. No male Prnp 0/0 mice were alive by 14 dpi with IAV/Aichi and IAV/ WSN (Fig 2D and 2E). However, about 90% and 75% of male WT mice survived at 14 dpi with IAV/Aichi and IAV/WSN, respectively (Fig 2D and 2E). Virus titers were higher in Prnp 0/0 lungs than in WT lungs after infection with IAV/Aichi and IAV/WSN (Fig 2D and 2E). Higher mortality was also observed in female Prnp 0/0 mice infected with IAV/Aichi and IAV/WSN, compared to control WT mice (Fig 2D and 2E). Taken together, these results indicate that Prnp 0/0 mice are highly susceptible to IAV infection, with higher mortality and higher virus loads in the lungs compared to WT mice, suggesting that PrP C could have a protective role against lethal infection with IAVs.

Transgenic expression of PrP C rescues Prnp 0/0 mice from IAV infection
To confirm that lack of PrP C is responsible for the higher susceptibility of Prnp 0/0 mice to IAV infection, 50 IFU of IAV/PR8 were intranasally infected into Tg(MoPrP)/Prnp 0/0 mice, in which multiple copies of the transgene encoding mouse PrP C are expressed on the Prnp 0/0 background [36]. Western blotting showed higher expression of PrP C in the lungs and brains of Tg(MoPrP)/Prnp 0/0 mice than in WT mice ( Fig 3A). Mortality was markedly reduced in male Tg(MoPrP)/Prnp 0/0 mice compared to male Prnp 0/0 littermates after infection ( Fig 3B). More than 90% of male Tg(MoPrP)/Prnp 0/0 mice survived the infection while only less than 10% of male Prnp 0/0 littermates remained alive at 14 dpi. Virus titers were also reduced in Tg (MoPrP)/Prnp 0/0 lungs compared to Prnp 0/0 lungs ( Fig 3B). A higher survival rate was also observed in female Tg(MoPrP)/Prnp 0/0 mice after infection, compared to control female Prnp 0/0 littermates ( Fig 3B). These results confirm that the higher susceptibility of Prnp 0/0 mice to IAV infection could result from the lack of PrP C .

The OR region is important for PrP C to protect against IAV infection
We then investigated whether the OR region might be involved in the protective role of PrP C against lethal infection with IAVs, by intranasal infection with 100 IFU of IAV/PR8 into Tg  transgenic mouse PrP with a deletion of the OR region alone on the Prnp 0/0 background [37]. Western blotting with 6D11 anti-PrP antibody, which recognizes residues 93-109 of mouse PrP, revealed expression of PrPΔOR in Tg(PrPΔOR)/Prnp 0/0 lungs and PrP C in WT lungs ( Fig  4A and 4B). SAF32 anti-PrP antibody, which recognizes the OR region, did not detect PrPΔOR in Tg(PrPΔOR)/Prnp 0/0 lungs (Fig 4A and 4B), confirming lack of the OR region in PrPΔOR. We increased the dose of IAV/PR8 for infection into Tg(PrPΔOR)/Prnp 0/0 mice to 100 IFU since they were highly resistant to 50 IFU of IAV/PR8. IAV/PR8 infection caused similar mortality in Tg(PrPΔOR)/Prnp 0/0 and Prnp 0/0 mice ( Fig 4C). However, mortality in these mice was significantly higher than that in control WT mice (Fig 4C). Only 10% of Tg (PrPΔOR)/Prnp 0/0 mice and no Prnp 0/0 mice survived the infection while 50% of WT mice were alive at 14 dpi ( Fig 4C). Tg(PrPΔOR)/Prnp 0/0 and Prnp 0/0 lungs showed similar virus titers, but they were higher than those in WT lungs ( Fig 4C). These results suggest that the OR region could play an important role for PrP C to protect against lethal infection with IAVs in mice.

More severe inflammation in IAV-infected Prnp 0/0 lungs
To gain insights into the protective role of PrP C against lethal infection with IAVs, we investigated the pathology of IAV/PR8 (50 IFU)-infected Prnp 0/0 and WT lungs. No macroscopic lesions were observed on the lung surface of control saline-administrated Prnp 0/0 and WT mice ( Fig 5A). In contrast, reddish lesions were evident on the surface of infected WT and Prnp 0/0 lungs at 5 and 8 dpi, with larger size and higher number of the lesions in the Prnp 0/0 lungs than in the WT lungs ( Fig 5A). Prnp 0/0 and WT lungs had increased wet weights after infection, with the Prnp 0/0 lungs being significantly heavier than the WT lungs ( Fig 5B), suggesting higher exudates in Prnp 0/0 lungs than in WT lungs after infection. Microscopic examinations showed higher infiltration of inflammatory cells in Prnp 0/0 lungs than in WT lungs after infection ( Fig 5C). Immunofluorescent staining also showed viral nucleocapsid protein NP accumulated in the inflammatory regions ( Fig 5C). Atelectatic areas were therefore larger in Prnp 0/0 lungs than in WT lungs after infection ( Fig 5D). We also investigated levels of inflammatory cytokines, including interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and interferon-γ (IFN-γ), in these infected lungs. All the cytokines examined had higher levels in Prnp 0/0 lungs than in WT lungs ( Fig 5E). We also investigated viral proteins in these infected lungs. Western blotting showed that viral proteins, including NP, NS1 nonstructural protein, and M2 matrix protein, became detectable in Prnp 0/0 and WT lungs at 3 dpi, reached a peak level at 5 dpi, and decreased at 8 dpi ( Fig 5F), with slightly but not significantly higher levels in the Prnp 0/0 lungs than in the WT lungs ( Fig 5G).

Higher epithelial cell damage in IAV-infected Prnp 0/0 lungs
To understand the protective mechanism of PrP C against lethal infection with IAVs, we investigated apoptotic cell death in IAV/PR8 (50 IFU)-infected Prnp 0/0 and WT lungs, by performing Western blotting for the cleaved fragments of the apoptotic marker caspase 3. Prnp 0/0 and WT lungs showed increased the fragments after infection ( Fig 6A). However, the fragments were higher in infected Prnp 0/0 lungs than in control WT lungs ( Fig 6A). Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) staining also showed more abundant apoptotic cells in alveolar and bronchiolar epithelial areas in Prnp 0/0 lungs than in WT lungs ( Fig 6B). We also performed Western blotting with anti-podoplanin, anti-SP-C, and anti-CC10 antibodies. Podoplanin levels were unaffected in Prnp 0/0 and WT lungs after infection ( Fig 6C). This is consistent with IAV/PR8 infection not damaging AT1 cells in C57BL/6 mice [38]. In contrast, SP-C and CC10 were markedly decreased in Prnp 0/0 and WT lungs after infection, with their levels significantly lower in infected Prnp 0/0 lungs than in control WT lungs ( Fig 6C). Consistently, immunofluorescence staining showed that SP-C-positive AT2 cells and CC10-positive Clara cells were less in infected Prnp 0/0 lungs than in control WT lungs, while podoplanin-positive AT1 cells were similarly observed in infected these lungs ( Fig  6D). These results suggest that AT2 and Clara cells in Prnp 0/0 lungs could be more vulnerable to apoptosis than those in WT lungs after infection with IAVs, and that PrP C could exert an anti-apoptotic activity in AT2 and Clara cells after infection with IAVs.

Anti-oxidant rescues Prnp 0/0 mice from lethal infection with IAV
To investigate the role of ROS in the higher mortality of IAVs-infected Prnp 0/0 mice, we addressed whether Prnp 0/0 mice could be rescued from lethal infection with IAV/PR8 by treatment with a ROS scavenger. To this end, we first measured ROS in the lungs of Prnp 0/0 and WT mice intranasally infected with IAV/PR8 (100 IFU). The virus dose used was higher than 1 MLD50 since it is assumed that the effect of a ROS scavenger on the survival rate of infected mice, if any, would be evaluated more easily for the mice developing mortality more than 50% after infection than those with less than 50% mortality after infection. No difference in ROS levels was detected between uninfected Prnp 0/0 and WT lungs (Fig 7A). IAV/PR8 infection at 5 dpi increased ROS levels in Prnp 0/0 and WT lungs ( Fig 7A). However, ROS levels were higher in Prnp 0/0 lungs than in control WT lungs (Fig 7A). ROS levels were also higher in infected Tg Protective role of prion protein from influenza virus and Prnp 0/0 mice uninfected (Un) and infected with 50 IFU of IAV/PR8 at 3, 5, and 8 dpi for podoplanin, SP-C, and CC10. Actb is an internal control. Right panels: Quantification of podoplanin, SP-C, and CC10 after normalization against β-actin (n = 3 for each mouse group). Signal (PrPΔOR)/Prnp 0/0 lungs than in control WT lungs at 5 dpi (Fig 7A). These results suggest that PrP C might exert an anti-oxidative activity to reduce ROS levels through the OR region in IAV-infected lungs. We then examined the anti-oxidative effect of butylated hydroxyanisole (BHA), a ROS scavenger, on ROS levels in the lungs of IAV/PR8 (100 IFU)-infected WT mice. Treatment with BHA for 4 days starting at 2 dpi effectively reduced ROS in infected WT lungs at 5 dpi (Fig 7B). We then similarly treated Prnp 0/0 and WT mice with BHA after intranasal infection with IAV/PR8 (100 IFU). The treatment decreased the mortality of infected WT mice (Fig 7C). This could be consistent with that ROS could be a major player in IAV infection-induced lung injury [30][31][32][33]. The mortality of infected Prnp 0/0 mice was also decreased to that of control WT mice after treatment with BHA ( Fig 7C). Viral titers were also decreased in Prnp 0/0 lungs to those in WT lungs after treatment with BHA ( Fig 7C). These results suggest that the higher ROS levels in Prnp 0/0 lungs could be involved in the higher mortality of Prnp 0/0 mice after infection with IAVs.

Higher expression of xanthine oxidase (XO) and lower activity of SOD1 in IAV-infected Prnp 0/0 lungs
To investigate the role of XO, a major ROS-generating enzyme in IAV-infected lungs [30], in IAV-infected Prnp 0/0 lungs, we first perform Western blotting of IAV/PR8 (100 IFU)-infected Prnp 0/0 and WT lungs for XO. Expression of XO was increased in these infected lungs ( Fig  8A). However, the expression levels of XO were higher in infected Prnp 0/0 lungs than in control WT lungs (Fig 8A). We then treated Prnp 0/0 and WT mice with the XO inhibitor allopurinol after intranasal infection with IAV/PR8 at 100 IFU, a dose higher than 1 MLD50 in WT mice. Allopurinol treatment starting from one day before intranasal infection with IAV/PR8 to 14 dpi reduced the mortality of Prnp 0/0 and WT mice to a similar rate (Fig 8B). Viral titers were also decreased in Prnp 0/0 lungs compared to those in WT lungs after treatment with allopurinol ( Fig 8B). These results suggest that XO could be a key ROS-generating enzyme in IAVinfected Prnp 0/0 and WT lungs, and that the higher expression of XO in IAV-infected Prnp 0/0 lungs could be involved in the higher mortality of IAVs-infected Prnp 0/0 mice probably through producing higher levels of ROS.
We also investigated infected Prnp 0/0 and WT lungs for the enzymatic activity of SOD, an anti-oxidative enzyme in IAV-infected lungs [31]. Western blotting revealed similar expression of SOD1 and SOD2 between uninfected and infected Prnp 0/0 or WT lungs (Fig 8C). However, the total SOD activity was significantly lower in uninfected Prnp 0/0 lungs than in uninfected WT lungs (Fig 8D). The SOD1-specific inhibitor diethyl-dithio-carbamate (DDC) reduced SOD activity in both uninfected Prnp 0/0 and WT lungs to the same levels (Fig 8D), suggesting that SOD1 activity might be impaired in Prnp 0/0 lungs. IAV/PR8 infection increased the total SOD activity in both WT and Prnp 0/0 lungs (Fig 8D). However, the activity was lower in infected Prnp 0/0 lungs than in control WT lungs (Fig 8D). DDC decreased the SOD activity in infected Prnp 0/0 and WT lungs to the levels in uninfected Prnp 0/0 lungs ( Fig  8D). These results suggest that SOD1 might not be fully activated in Prnp 0/0 lungs after IAV infection. Lower SOD1 activity was also detected in infected and uninfected Tg(PrPΔOR)/ Prnp 0/0 lungs than in control WT lungs (Fig 8D). Cu ions, which are important for SOD1 activity, were lower in uninfected Prnp 0/0 and Tg(PrPΔOR)/Prnp 0/0 lungs than in control WT intensity for podoplanin, SP-C, and CC10 in Prnp 0/0 lungs was evaluated against that in uninfected WT lungs. (D) Immunofluorescence staining of uninfected and IAV/PR8 (50 IFU)-infected WT and Prnp 0/0 lungs at 5 dpi with antibodies against podoplanin, SP-C and CC10, and with DAPI. Right panels: Immunofluorescent staining for the viral protein NP in the lungs of these mice. Bar, 400 μm. NS, not significant; Ã , p<0.05; ÃÃ , p<0.01. Error bars, SD. https://doi.org/10.1371/journal.ppat.1007049.g006 Protective role of prion protein from influenza virus lungs (Fig 8E). These results suggest that PrP C might function to maintain Cu levels and thereby might regulate SOD1 activity through the Cu-binding OR region in lungs.

Primary Prnp 0/0 lung cells are susceptible to IAV infection
To further gain insights into the protective role of PrP C in IAV infection, we infected primary lung cells from WT, Prnp 0/0 and Tg(MoPrP)/Prnp 0/0 mice with IAV/PR8 at 1.0 multiplicity of

Lipopolysaccharide (LPS) induces similar injuries in Prnp 0/0 and WT lungs
Intranasal administration of LPS is known to cause lung injuries in mice [39]. To investigate whether PrP C might be also protective against LPS-induced lung injuries, we intranasally administrated LPS into

Discussion
In the present study, we showed that Prnp 0/0 mice were highly susceptible to infection with IAVs, with markedly higher mortality, compared to control WT mice. Pathological changes were more severe, inflammatory cytokines including IL-6, TNF-α, and IFN-γ were higher, and viral loads were higher in IAV/PR8-infected Prnp 0/0 lungs. We confirmed that the higher mortality of infected Prnp 0/0 mice is due to lack of PrP C , by demonstrating that transgenic expression of mouse PrP C rescued Prnp 0/0 mice from lethal infection with IAV/PR8. We also showed that mouse PrP lacking the OR region failed to protect Prnp 0/0 mice from the lethal infection with IAV/PR8. These results suggest that PrP C could have a protective role against lethal infection with IAVs through the OR region in mice.
Prnp 0/0 mice activated innate and adaptive immune responses against IAV infection. These results rule out the possibility that lack of PrP C might cause defective immune responses against IAV infection, therefore Prnp 0/0 mice being highly susceptible to IAV infection. PrP C was expressed in alveolar AT1 and 2 epithelial cells and bronchiolar Clara epithelial cells in lungs. Other investigators also reported expression of PrP C in alveolar walls and Clara cells [8,40]. Consistent with the previous report showing that AT1 cells were resistant to infection with IAV/PR8 in WT mice [38], AT1 cells were unaffected by infection with IAV/PR8 not only in WT lungs but also in Prnp 0/0 lungs. In contrast, infection with IAV/PR8 markedly damaged AT2 and Clara cells in Prnp 0/0 and WT lungs. However, these epithelial cells in Prnp 0/0 lungs were more susceptible to the infection than those in WT lungs. Caspase 3 was activated more robustly in Prnp 0/0 lungs than in WT lungs after infection with IAV/PR8. TUNEL staining also displayed more abundant apoptotic cells in the alveolar and bronchiolar epithelial areas of infected Prnp 0/0 lungs than in control WT lungs. Primary Prnp 0/0 lung culture cells were also vulnerable to IAV/PR8 infection-induced apoptosis compared to control WT lung cells. These results suggest that AT2 and Clara epithelial cells in Prnp 0/0 lungs might be more vulnerable to IAV infection-induced apoptosis than those in WT lungs, and that PrP C might play an anti-apoptotic role in these lung epithelial cells in a cell-autonomous way. However, intranasal administration with LPS, which induces lung injuries through binding to Tolllike receptor 4 (TLR4) [41], similarly activated caspase 3 in Prnp 0/0 and WT lungs, suggesting that the anti-apoptotic activity of PrP C has no effect on the LPS/TLR4-induced apoptosis in lungs.
Viral loads were significantly but only slightly higher in Prnp 0/0 lungs than in WT lungs after infection with IAV/PR8. It has been shown that caspase 3 activation induces efficient replication of IAV in cells [42]. It is thus possible that the slightly higher viral loads in IAV/ PR8-infected Prnp 0/0 lungs might be associated with the higher activation of caspase 3 observed in the lungs. Thus, PrP C might exert its protective activity against IAV infection through its anti-apoptotic activity in lung epithelial cells, not through directly affecting viral replication efficiency in lungs. However, the possibility remains unanswered if PrP C could directly affect IAV replication in the lungs, thereby reducing viral loads and eventually repressing caspase 3 activation in the lungs.
AT2 cells are small cuboidal cells covering about 2-5% of the alveolar surface area and secreting surfactant proteins, which are important to reduce alveolar surface tension [43,44]. Clara cells are the predominant cell type in bronchioles and known as important progenitor cells for the repair of bronchiolar epithelia [45]. Recently, it was reported that Clara cells are also major progenitor cells for alveolar epithelial regeneration through differentiation to AT1 and 2 alveolar cells after IAV infection [46,47]. AT2 and Clara cells were more severely damaged in Prnp 0/0 lungs than in WT lungs after infection with IAV/PR8. It is thus possible that the AT2 cells-mediated regulation of alveolar surface tension and the Clara cells-mediated alveolar and bronchiolar epithelia regeneration after IAV infection might be disturbed more severely in Prnp 0/0 lungs than in WT lungs after infection with IAVs, eventually causing higher mortality in Prnp 0/0 mice infected with IAVs.
We showed that ROS levels were higher in IAV/PR8-infected Prnp 0/0 lungs than in control WT lungs. We also showed that the ROS scavenger BHA rescued Prnp 0/0 mice from lethal infection with IAV/PR8, reducing mortality to the levels in IAV/PR8-infected, BHA-treated control WT mice, suggesting that the higher ROS levels in infected Prnp 0/0 lungs could be responsible for the higher mortality of Prnp 0/0 mice infected with IAVs. It has been shown that Prnp 0/0 cells were more susceptible to treatment with agents inducing oxidative stress, readily succumbing to apoptosis, compared with WT cells [23,48], suggesting that PrP C could have a protective role against oxidative stress-induced apoptosis. Therefore, PrP C might play an antioxidative role in lungs after infection with IAVs, thereby reducing ROS levels and protecting lung epithelial cells from IAV infection-induced apoptosis. We also demonstrated higher ROS levels in Tg(PrPΔOR)/Prnp 0/0 lungs than in WT lungs after infection with IAV/PR8, suggesting that the OR region could be important for PrP C to exert the anti-oxidative activity in lungs after infection with IAVs.
XO was shown to be a major ROS-generating enzyme in IAV-infected lungs [30]. We showed that XO expression was elevated in infected Prnp 0/0 lungs compared to control WT lungs. We also showed that the XO inhibitor allopurinol rescued Prnp 0/0 mice from lethal infection with IAV/PR8. These results suggest that the XO up-regulation observed in infected Prnp 0/0 lungs might be responsible for the higher mortality in Prnp 0/0 mice infected with IAVs. It has been shown that inflammatory cytokines such as TNF-α and IFN-γ up-regulate the expression of XO [49][50][51]. Higher levels of these cytokines were detected in infected Prnp 0/0 lungs than in control WT lungs, suggesting that the higher expression of XO in infected Prnp 0/ 0 lungs might be induced by the higher levels of these cytokines in the lungs.
We also showed that IAV/PR8 infection increased SOD1 activity in Prnp 0/0 and WT lungs. However, SOD1 was not fully activated in infected Prnp 0/0 lungs compared to control WT lungs, suggesting that SOD1 activation might be disturbed in Prnp 0/0 lungs infected with IAVs. Together with the reported results that administration of pyran polymer-conjugated SOD1 successfully reduced the mortality of WT mice infected with IAV [31], it is suggested that the lower activity of SOD1 in infected Prnp 0/0 lungs might be responsible for the higher mortality of Prnp 0/0 mice infected with IAVs. Cu ions are important for the SOD1 enzymatic activity. We found that Cu ion content were lower in Prnp 0/0 lungs than in WT lungs, suggesting that the lower activity of SOD1 in Prnp 0/0 lungs might be due to the lower Cu content in the lungs. Reduced SOD1 activity and lower Cu content were also detected in Tg(PrPΔOR)/Prnp 0/0 lungs. It is thus possible that PrP C might have a role to maintain Cu levels in lungs through the OR region, thereby regulating SOD1 activity and eventually exerting an anti-oxidative activity in lungs. PrP C is known to bind to Cu ions via the OR region, suggesting that PrP C might transfer the bound Cu ions to and activate SOD1 [23,24]. However, the exact mechanism of how PrP C is involved in the activation of SOD1 remains to be determined. It has been also proposed that PrP C itself could have SOD activity [52]. However, other investigators failed to confirm this proposed SOD activity in PrP C [53,54]. Elucidation of the mechanism underlying the anti-oxidative function of PrP C could be helpful for further understanding the pathogenesis of IAV infection and for development of anti-influenza therapeutics based on the PrP C -mediated protective mechanism.
Anti-oxidative therapeutics against IAV infection, by targeting the ROS-generating enzymes or by administrating anti-oxidants or anti-oxidant enzymes, has been shown to successfully protect mice from lethal infection with IAVs [30][31][32][33]. Our current results showing that PrP C could have a protective role against lethal infection with IAVs in mice possible by exerting ant-oxidative activity, suggest PrP C to be a new target molecule for anti-oxidative therapeutics against IAV infection. It has been reported that PrP C elicited a protective signal against anisomycin-induced apoptosis in neurons via interaction with stress-inducible protein 1 (STI1), a STI1-derived peptide, or anti-PrP antibodies [55,56], and that the interaction with STI1 could be involved in PrP C -dependent activation of SOD [57]. It is thus interesting to investigate whether these ligands could elicit the protective activity of PrP C against IAV infection.

Ethics statement
All animal experiments were conducted in compliance with Japanese legislation (Act on Welfare and Management of Animals). The Ethics Committee of Animal Care and Experimentation of Tokushima University approved the animal experiments in this study (approval number T27-86). Animals were cared for in accordance with The Guiding Principle for Animal Care and Experimentation of Tokushima University and guidelines under the jurisdiction of the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Virus preparation
IAV A/PR/8/34 (H1N1), A/Aichi/2/68 (H3N2), and A/WSN/33 (H1N1) were injected into the allantoic sac of 11-day-old chicken embryos in eggs and incubated at 36˚C for 48 h. The eggs were chilled at 4˚C for at least for 4 h prior to harvesting the allantoic fluids. Cellular debris in the allantoic fluids was removed by centrifugation at 2,380×g at 4˚C for 30 min. The clarified allantoic fluids were layered over a 20% sucrose cushion and centrifuged at 25,000×g at 4˚C for 120 min. The pellet containing viruses was suspended in phosphate-buffered saline (PBS), and stored in multiple aliquots at -80˚C until used.

Intranasal infection with IAVs
Male and female mice aged 5 weeks were intranasally inoculated with IAVs in a total volume of 20 μL (10 μL in each nasal cavity), and monitored for survival and weight loss for 14 days. The IAV stock aliquot was thawed and diluted in saline before used.

Intranasal administration with LPS
Sixty micrograms of LPS (026:B6, Sigma-Aldrich, St Louis, MO) in 20 μL PBS were intranasally administered into a mouse using a micropipette (10 μL in each nasal cavity). PBS was similarly administered as a control.

Treatment with BHA and allopurinol
BHA and allopurinol were purchased from Sigma-Aldrich. BHA was dissolved in dimethyl sulfoxide (DMSO) and stored at -20˚C until use. BHA was orally administered at 200 mg/kg/ day using a sonde needle from 2 to 5 dpi. According to the Material Safety Data Sheet (MSDS) (ScienceLab. com, Inc. Dickinson, Texas), the oral LD50 of BHA is 1,100 mg/kg in mice. Allopurinol was prepared in PBS and then stored at -20˚C until use. Allopurinol was orally administered at 2 mg/kg/day using a sonde needle from -1 to 14 dpi. According to the MSDS (Sigma-Aldrich), the oral LD50 of allopurinol is 78 mg/kg in mice.

Tissue and cell homogenization
Tissues were homogenized using a Polytron homogenizer (PT 2100, Brinkman Instruments, Inc., Westbury, NY) in 2 mL of PBS for measurement of cytokines and Cu ions, and in a lysis buffer (0.5% Triton X-100, 0.5% sodium deoxycholate, 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM EDTA) containing protease inhibitor cocktail (Nakalai Tesque, Kyoto, Japan) for Western blotting and for measurement of ROS levels and SOD activity. Cells were homogenized in a protease inhibitor cocktail (Nakalai Tesque)-containing lysis buffer and subjected to Western blotting and measurement of ROS levels and SOD activity. The homogenates were clarified by centrifugation at 1,000×g for 2 min at 4˚C. Protein concentration of the homogenates was measured using the BCA method (Thermo Scientific, Rockford, IL).

Determination of virus titers
Virus titers were expressed as IFU/mL. IFU/mL was determined using Madin-Darby canine kidney (MDCK) cells as follows. MDCK monolayer cells were incubated with 10-fold serial dilutions of each sample of interest for 14 h at 37˚C. The cells were then fixed with 4% paraformaldehyde, permeabilized with 0.3% Triton-X100 in PBS, and immunostained with anti-NP monoclonal antibodies (GeneTex, Irvine, CA). Signals were visualized using horseradish peroxidase (HRP)-conjugated anti-rabbit IgG antibodies (GE Healthcare, Waukesha, WI) and True Blue Peroxidase Substrate (KPL, Gaithersburg, MD). IFU/mL was defined as the number of the cells positive for the anti-NP signals in 1 mL of each sample.

Hematoxylin-eosin (H-E) staining
After euthanasia of mice, lungs were quickly removed, fixed with 4% paraformaldehyde, dehydrated, embedded in paraffin, and sliced into 5 μm-thick tissue sections. The sections were deparaffinized, rehydrated, and stained with hematoxylin for 5 min and eosin for 30 sec.

Determination of atelectatic lung areas
The atelectatic lung area was evaluated using Photoshop software (Adobe, San Jose, CA) and ImageJ software (NIH, Bethesda, MD). Briefly, the original RGB color images of H-E stained lung sections were converted to black-on-white images using Photoshop software and saved in TIFF format. The binary images in TIFF format were again converted into a white-on-black image using the ImageJ application. Atelectatic lung area was expressed as the area of white pixels, which represent solid areas, against total lung area (white and black pixels).

TUNEL staining
TUNEL staining was performed using the in situ cell death detection kit and fluorescein (Roche Diagnostics, Mannheim, Germany) in accordance with the manufacturer's protocol. In brief, the deparaffinized tissue sections were treated with 20 μg/mL proteinase K in 10 mM Tris-HCl for 30 min at room temperature and incubated in the TUNEL reaction mixture for 1 h at 37˚C in a humidified dark chamber. The sections were washed with PBS for 5 min 3 times and signals were then detected using BIOREVO BZ-9000 (Keyence).

Western blotting
Proteins in each sample were denatured by boiling for 5 min in Laemmli's sample buffer and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were electrically transferred onto Immobilon-PVDF membranes (Millipore, Bedford, MA), and membranes were blocked for 2 h with 5% non-fat dry milk-containing TBST (0.1% Tween-20, 100 mM NaCl, 10 mM Tris-HCl, ph 7.

Enzyme-linked immunosorbent assay (ELISA) for cytokines
IL-6, TNF-α, and IFN-γ levels in samples were determined using a Quantikine ELISA kit (R&D systems) according to the respective protocols provided by the manufacturer. In brief, the samples were diluted 1:1 with the assay diluent provided in the kit and added to the ELISA microplate wells. The protein standards for IL-6, TNF-α, and IFN-γ were also added to other wells. The plates were then left for 2 h at room temperature, and the wells were washed with wash buffer 5 times and mouse IL-6, TNF-α, or IFN-γ conjugate added followed by incubation for 2 h. The wells were then washed with the wash buffer and the substrate reagent added followed by incubation for 30 min. The reaction was stopped by addition of the stop solution. The optical density of each well was measured at 450 nm in an automated microplate reader (Thermo LabSystems, MA, USA). The amounts of IL-6, TNF-α, or IFN-γ in each sample were determined using the standard curve for the amounts of IL-6, TNF-α, or IFN-γ.

ROS measurement
ROS concentration in samples was measured using an OxiSelect Intracellular ROS Assay Kit (Cell Biolabs, San Diego, CA). The assay uses 2',7'-dichlorodihydrofluorescin diacetate (DCFH-DA), which is deacetylated to non-fluorescent 2',7'-dichlorodihydrofluorescin and then oxidized by ROS to highly fluorescent 2',7'-dichlorofluorescin (DCF). Each of the samples were mixed with 1×DCFH-DA solution in a 96-well black plate and incubated at 37˚C for 48 h. ROS concentration in the samples was measured by determining the fluorescence intensities of DCF at 480 nm using Spectra Max Gemini EM (Molecular devices, Sunnyvale, CA).

Measurement of SOD activity
SOD activity in samples was determined using an OxiSelect Superoxide dismutase activity assay kit (Cell Biolabs). This assay uses a xanthine/XO system to produce superoxide anions, which reduce chromagen to produce a formazan dye, which is colorimetrically detectable at 490 nm. SOD activity in the samples was determined as the inhibition of formazan dye production. Each of the samples was mixed with 1× XO solution in a 96-well black plate and incubated at 37˚C for 60 min and the formazan dye produced was colorimetrically detected at 490 nm using Spectra Max Plus (Molecular devices). SOD1 inhibition was achieved by the addition of DDC (Sigma-Aldrich) to a final concentration of 1 mM into the mixture as described elsewhere [59][60][61].

Measurement of Cu ions
Total copper levels in samples were assessed using the Metallo assay low copper LS kit (Metallogenics, Chiba, Japan) according to the manufacturer's instructions. The pH of the samples was adjusted to 3.0 by adding a small amount of 0.25 mM HCl. Color reagent was then added and incubated at room temperature for 10 min. The copper concentration in the samples was calculated by measuring the absorbance at 580 nm using Spectra Max Plus (Molecular devices).

Preparation of primary lung cell culture and viral infection
After euthanasia of mice, whole lungs were removed after perfusion of the mice with saline. The lungs were then cut into pieces and sieved through a 40 μm nylon cell strainer (BD Falcon, Franklin Lakes, NJ) with PBS. Lung cells were then collected by centrifugation at 1,000×g at 4˚C for 2 min. The collected cells were suspended in Ham's F-12K medium (Life Technologies, Grand Island, NY) supplemented with 15% FBS and cultured for 24 h. The cells were then cultured in F-12K medium without FBS in a 96-well plate at a density of 5.0×10 4 cells/well for another 24 h, and infected with IAV/PR8 at a 1 MOI in the presence of 0.05% trypsin (Invitrogen). Cell viability was assessed using a Cell Counting Kit-8 (Dojindo).

Knockdown of PrP C expression and IAV infection in A549 cells
Human lung epithelial A549 cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Wako Pure Chemicals) with 10% FBS and transfected with non-targeting control siRNA (cat: D-001210-01-05, Thermo Scientific) and human PrP-specific siRNA (cat: D-011101-02, Thermo Scientific). Briefly, 6.25 μL of RNAiMAX transfection reagent (Invitrogen) was mixed with 125 μL of Opti-MEM (Life Technologies) and incubated for 5 min at room temperature. In a separate tube, siRNA was added to 125 μL of Opti-MEM at a final concentration of 150 nM and the solution was then mixed with the RNAiMAX mixture for 20 min at room temperature. The siRNA/RNAiMAX mixture was then added to A549 cells in a 6-well plate. At 24 h after transfection, cells were washed with PBS and infected with IAV/PR8 at 1 MOI in 10% FBS-containing DMEM. Cells were collected, lysed, and subjected to Western blot analysis 24 h after infection.

Preparation of splenocytes
After euthanasia of mice, whole spleens were removed and sieved through a 40 μm nylon cell strainer (BD Falcon) with PBS. Splenocytes were then harvested by centrifugation at 1,000×g for 2 min at 4˚C. The resulting pellet was suspended in ACK buffer (0.15 M NH 4 Cl, 1.0 mM KHCO 3 , 0.1 mM Na 2 EDTA, pH 7.2) at room temperature for 2 min to disrupt red blood cells and centrifuged at 1,000×g for 2 min at 4˚C. The collected splenocytes were adjusted to a concentration of 2.5×10 6 cells/mL in RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% FBS, 1% L-glutamine (Sigma-Aldrich), 2 μM L-glutamate (Sigma-Aldrich), non-essential amino acids (Sigma-Aldrich), 10 mM HEPES, and 1 mM sodium pyruvate.

RT-PCR
RT-PCR was performed using OneStep RT-PCR Kit (QIAGEN, Hilden, Germany) according to manufacturer's recommendations. Total RNA was first extracted from tissues using RNeasy Mini Kit (QIAGEN). Tissue homogenates in buffer RLT were transferred to a QIAshredder spin column (QIAGEN). The flow-through was mixed with 1 volume of 70% ethanol and then transfer to an RNeasy spin column (QIAGEN). Total RNA bound to the membrane was washed with buffer RW1 and then with buffer RPE, and eluted with RNase-free water. Eight ng of total RNA was then mixed with primers, dNTPs and OneStep RT-PCR enzyme mix. The mixture was incubated at 50˚C for 30 min at RT and then subjected to PCR reaction (Initial PCR activation step at 95˚C for 15 min; 3-step cycling: Denaturation at 94˚C for 30 sec, Annealing at 56˚C for 30 sec, Extension at 72˚C for 1 min; Final extension at 72˚C for 10 min). Sequences of the primers used and the number of PCR cycles used for each gene examined are given in S1 Table. The products were analyzed by 2% agarose gel electrophoresis.

Determination of IAV/PR8-specific IgG and IgM titers
IAV/PR8-specific IgG and IgM titers in plasma were determined by ELISA. Each well of a 96 well immunoplate (Thermo Fisher Scientific, Roskilde, Denmark) were coated with the already prepared split IAV/PR8 vaccine [62] in PBS overnight at 4˚C. The wells were then washed with PBS 3 times and blocked with PBS containing 4% Block Ace (Megmilk Snow Brand Co., Ltd., Hokkaido, Japan) for 1 h at 37˚C. Mouse plasma samples were first diluted to 1:16 and subsequently 1:2 in PBS and added to the wells at 37˚C for 4 h. The wells were washed with PBS containing 0.05% Tween-20 3 times, and immune complexes were detected using HRPconjugated goat anti-mouse IgM or IgG antibodies (Bethyl Laboratories, Inc., Montgomery, TX) and 1-Step Ultra TMB-ELISA (Thermo Scientific). The signals were detected spectrophotometrically at 450 nm using Spectra Max Plus (Molecular devices). Antibody titers are defined as the reciprocal of the highest dilution of samples for which the optical density was at least twice of that of the negative control samples, and are expressed as reciprocal log 2 titers.

Statistical analysis
Survival rates were analyzed using the log-rank test. All other data were analyzed using the Student's t-test.
Supporting information S1 Table. List of genes, sequences of the primers, and the number of cycles used for RT-PCR gene expression analysis.