Interferon kappa (IFN-κ) is a member of the type I interferons (IFN-I), first reported to inhibit encephalomyocarditis virus (EMCV) virus replication in epidermal keratinocytes in 2001. Further studies have shown that IFN-κ is constitutively expressed in human keratinocytes at mucosal surfaces and has anti-human papillomavirus and anti-herpes simplex virus 1 activity.[2,3]IFN-κ seems to be a relatively ancient and evolutionally conservative host defense factor, and it has been shown that chicken IFN-κ markedly inhibits the replication of avian RNA viruses in vivo. In addition, bovine IFN-κ also exhibits antiviral activity by activating multiple interferon-stimulated genes (ISGs), such as ISG15, ISG56, and MX1. Poultry IFN-κ was constitutively expressed in keratinocytes and the spleen, skin, lung, and peripheral blood mononuclear cells (PBMCs) was significantly induced in PBMCs after viral infection. Recently, two clinical studies have shown that treatment with aerosol inhalation of IFN-κ plus trefoil factor 2 (TFF2) could facilitate clinical improvement in patients with moderate COVID-19. Among them, IFN-κ inhibits viral replication, thereby facilitating viral clearance.[7,8]
Although IFN-κ is classified as IFN-I, it shares only ∼30% homology with other family members. Previous studies have indicated that IFN-κ utilizes the same receptor as other IFN-Is to activate the classic JAK-STAT signaling pathway.[1,5] However, accumulating evidence has shown that IFN-κ is different from other IFN-Is. IFN-κ exhibited a lower binding affinity for cell-surface IFNAR2 than IFN-α2 and IFN-ω. In addition, a recent study revealed that IFN-κ suppressed the replication of influenza A viruses through the IFNAR-MAPK-Fos-CHD6 axis, which is different from the classical JAK-STAT pathway induced by other IFN-Is. Furthermore, IFN-κ displayed cell-associated antiviral activity against EMCV and hepatitis C virus, which has not been observed for other IFN-I, indicating that IFN-κ is a unique IFN-I and may be highly different from other IFN-Is.
This study further analyzed the structural and functional differences between IFN-κ and IFN-I. We found that human IFN-κ upregulated ISGs and inhibited influenza virus replication in a membrane-anchored manner in eukaryotic cells. Expression, location, and function of IFN-κ were associated with the 1-37aa at the N-terminus of IFN-κ. In addition, soluble mature human IFN-κ purified from a prokaryotic expression system restricts RNA virus replication by inducing ISGs. Furthermore, IFN-κ induces ISG response faster but with less intensity and a shorter half-life than the response of IFN-α2. Since IFN-κ is usually constitutively expressed in human keratinocytes and mucosal epithelial cells, it is rationalized that IFN-κ is likely to function as the first wave of attack against viral invasion, followed by stronger and prolonged strikes from other interferons, such as IFN-α.
Materials and methods
Full-length human IFN-κ with a 6xHis tag at the C terminus was amplified and cloned into the eukaryotic expression vector pSV1.0, while other mutants including IFN-κ with a deletion of 1-27aa (signal peptide), 1-37aa, 28-37 aa at the N terminus, or with a replaced TFF2 signal peptide, were constructed with ClonExpress II One Step Cloning Kit (C112-01, Vazyme.com) according to the manufacturer's instructions. In addition, green fluorescent protein (GFP), mouse IFN-κ-6xHis, and human/ mouse IFN-α2 were also cloned into the pSV1.0 vector.
Western blot (WB)
Whole-cell lysates were prepared and separated by denaturing SDS-PAGE gel electrophoresis, followed by transfer to polyvinylidene fluoride (PVDF) membranes. Then, the membranes were blocked with 5% dried skim milk in PBS supplemented with 0.05% Tween 20 (PBST) for 1hour at room temperature. Next, the membranes were incubated with the indicated primary antibodies overnight at 4°C and horseradish peroxidase-conjugated secondary antibodies against mouse (1:3000, 33201ES60, Yeasen) or rabbit (1:3000, 33101ES60, Yeasen) for 1hour at 20°C to 30°C. Finally, the protein bands on the membranes were visualized using the Odyssey®Fc Imaging System (LI-COR Biotechnology). The primary antibodies used included anti-NP (1:5000, GeneTex, GTX125989), anti-M1 (1:1000, GeneTex, GTX125928), anti-IFN-κ (1:1000, Abnova, H00056832-M01), anti-His (1:1000, ZSGB-Bio, TA-02), anti-Interferon Inducible Transmembrane Protein 3 (IFITM3) (1:1000, Proteintech, 11714-1-AP), anti-MX1 (1:1000, Proteintech, 13750-1-AP), and anti-β-actin (1:5000, ABclonal, AC004).
Construction of stable cell line
Human IFN-κ was cloned into pHAGE vector carrying GFP, and pHAGE (negative control) or pHAGE-IFN-κ combined with pSPAX and pMG2D were co-transfected into 293T cells and packaged into lentivirus. Then, 293FT cells were infected with lentivirus for 2 hours, and 48 hours later, GFP-positive 293FT cells were sorted by flow cytometry for culture. After sorting, the positive rate of GFP was more than 95%.
Cytoplasm-membraneproteinseparation was carriedoutusingthe Mem-PERTM Plus Membrane Protein Extraction Kit (89842, ThermoFisherScientific) accordingtothe manufacturer'sprotocol.
293FT cells stably expressing IFN-κ were placed in a 24-well plate in the lower chamber, and A549 cells were placed in Transwell inserts (Corning Costar, Cambridge, MA, USA) in the upper chamber. They were separated with membranes of 0.4 μm pore size for co-culture, allowing only supernatants to pass through, not cells. After 24 hours, supernatants and A549 cells were collected and subjected to WB analysis.
Cell contact and sorting
For the cell-to-cell contact interaction assay, A549 cells were plated in a 24-well plate for 48 hours to allow A549 cells to attach to the wells, and then control or IFN-κ stably expressing 293FT cells were added to the wells. After incubation for 30 minutes, control or IFN-κ stably expressed 293FT cells were rapidly removed from the wells. A549 cells were washed twice with PBS and cultured for another 24 hours. Then, A549 cells were lysed and subjected to WB analysis.
For cell sorting, A549 cells were labeled with eFluor 450 dye and then immediately mixed with control or IFN-κ stably expressing 293FT cells (carrying GFP). After co-culture for 40 hours, A549 and 293FT cells were sorted by flow cytometry according to their different fluorescence and subjected to WB analysis.
The human IFN-κ cDNA sequence (without signal peptide sequence) was chemically synthesized according to the optimized codons of Escherichia coli and then expressed in E. coli strain BL21(DE3). IFN-κ was expressed as inclusion bodies and purified according to a previously described method. The activity of purified IFN-κ protein was determined as 1.16 × 106 U/ mg based on the WISH-VSV method from Novoprotein (China).
For influenza virus infection, A549 cells were cultured in 24-well plates and treated with IFN protein or supernatant for the indicated time. Then, 5 multiplicity of infection (moi) of PR8 or H9N2 virus was added to the cells in a volume of 200 μL Dulbecco's Modified Eagle Medium (DMEM) supplemented with diethylaminoethyl (DEAE)-dextran alone. After infection for 2 hours, virus was discarded and cells were cultured with DMEM maintenance culture medium (2% FBS plus 1%PS) for 48 hours. For SARS-CoV-2 infection, Huh7 cells were pretreated with 1000U/mL IFN-κ or IFN-α2 for 24 hours, and infected with 0.1 moi SARS-CoV-2 for 2 hours. Then, the virus was discarded and the cells were cultured with DMEM maintenance culture medium. At 24, 48, and 72 hours post-infection, cells and supernatants were collected.
Quantitative polymerase chain reaction (qPCR)
Cells were treated with TRIzol (15596018, Life Technologies), and total RNA was extracted using Direct-zolTM RNA MiniPrep (R2052, Zymo Research) according to the manufacturer's instructions. Viral RNA in the supernatants was extracted using the QIAamp Viral RNA Mini Kit (52904, Qiagen, Germany). The viral load of SARS-CoV-2 was analyzed by quantitative Reverse transcription polymerase chain reaction (RT-PCR), and the procedures were carried out according to the instructions of the One-Step RT-PCR kit (210212, Qiagen). Primers F: 5′-ACTTCTTTTTCTTGCTTTCGTGGT-3’, R: 5’-GCAGCAG-TACGC ACACAATC-3’, and probe 5 Cy5-CTAGTTACAC-TAGCCTCCTTACTGC-3BHQ2 were used for quantitative RT-PCR amplification, and the standard was diluted 10 times with pUC57-2019-nCoV-E plasmid (initial concentration: 109 copies/μL). The thermal cycling conditions were as follows: 42°C for 15 minutes, followed by 94°C for 1 minute, and 45 cycles of 95°C for 15 seconds and 60°C for 30 seconds. In addition, mRNA levels were analyzed with quantitative RT-PCR (SYBR Green method) on a Real-Time PCR System (Eppendorf 7500) using the following primers: IFITM3 F: 5′-TCGTCTGGTCCCTGTTCA ACA-3′; R: 5′-TCCT GTCCCTAGACTTCACGGA-3’; ISG15 F: 5′-GGTGGACAAATGCGACGAAC-3′; R: 5′-ATGCTGGTG-GAGGCCCT TAG-3′; MX1 F:5′-GTGCATTGCAGAAGGT-CAGA-3′; R: 5′-CGGCTAACGGAT AAGCAGAG-3′; OAS3 F: 5′-ACAGCTGAAAGCCTTTTGGA-3′; R: 5′-GCATTAAA GG CAGGAAGCAC-3′. GAPDH F: 5′-ACGGATTTGG TCGTAT TGGG-3′; R: 5′-ATCTCGCTCCTGG AAGATGG-3’.
After treatment with 1000U/mL IFN-κ or IFN-α2 for the indicated time, A549 cells were harvested and total RNA was extracted to construct a cDNA library (BWR001, BioWavelet). Then, PCR was amplified using P5/P7adapter primers (BWR002, BioWavelet). All libraries were sequenced using the Illumina HiSeq platform. Normalized data were used to analyze the expression of ISGs.
Statistical analyses were conducted using the GraphPad software (Prism 5, San Diego, CA, USA). Data were expressed as mean± standard error of mean (SEM). A t-test was applied when comparing two groups. A P-value of ≤0.05 was considered statistically significant.
IFN-κ on cell membrane induced the expression of IFITM3, but not IFN-κ in supernatants
A previous study indicated that the antiviral activity of transiently expressed human IFN-κ is cell-associated. To further characterize the expression and function of IFN-κ in cells and supernatants, we first constructed a 293FT cell line stably expressing IFN-κ with a 6xHis tag. It was found that IFN-κ was detected both in cells and supernatants, but not in supernatants of the negative control (nc) 293FT cell line [Figure 1A]. Next step, we used supernatants from nc-293FT or IFN-κ-293FT cell lines in various proportions to stimulate A549 cells for 24 hours. Surprisingly, no supernatants induced the expression of IFITM3 except for the recombinant IFN-Is [Figure 1B]. However, the expression of IFITM3 was upregulated in the IFN-κ-293FT cell line but not in the nc-293FT cell line [Figure 1C]. In addition, supernatant transfer of human IFN-κ did not inhibit PR8 replication in A549 and L929 cells (Supplementary Figure 1a–c, https://links.lww.com/IDI/A9). However, the PR8 virus was infected and replicated in the nc-293FT cell line, but not in the IFN-κ-293FT cell line (Supplementary Figure 1d, https://links.lww.com/IDI/A9). These data showed that although IFN-κ was also expressed in the supernatants, it did not induce ISGs and inhibited influenza virus replication. In contrast, IFN-κ in cells restrained the influenza virus, an effect that was correlated with upregulation of IFITM3. Next, to further investigate the location of IFN-κ in cells, the cell membrane and cytoplasm were separated, and IFN-κ was determined in these two components. As shown in Figure 1D, IFN-κ is mainly located on the cell membrane.
To verify this result, we measured the activity of IFN-κ in the supernatants using the Transwell system. This system allows the continuous free flow of medium between the two cultures without interaction between cells in the two chambers. Naive A549 cells were placed in the upper transwell chamber, and IFN-κ-293FT cells were placed in the lower cell culture plate. The cells were cultured in free medium for 24 hours. The results showed that IFN-κ-293FT cells expressed IFITM3, whereas A549 cells did not induce the expression of IFITM3 [Figure 1E], verifying that IFN-κ in supernatants was insufficient to stimulate the ISG response. To further validate that IFN-κ on the cell membrane could induce ISGs, we carried out cell-cell surface contact assays. A549 cells were found to be stimulated to express IFITM3 after co-culture for 30 minute with the IFN-κ-293FT cell line, but not with the nc-293FT cell line [Figure 1F]. In addition, when 293FT and A549 cells were mixed for 40 hours and sorted to detect IFITM3, the results indicated that the expression of IFITM3 was upregulated only in A549 cells mixed with the IFN-κ-293FT cell line, but not in the nc-293FT cell line [Figure 1G]. These results demonstrated that IFN-κ was mainly localized on the cell membrane and could effectively induce IFITM3 expression either in IFN-κ-expressing cells (auto-stimulation) or in adjacent cells via cell-cell direct contact (para-stimulation).
Structure and function analysis of human IFN-κ
To clarify how IFN-κ was located on the cell membrane, we first predicted the potential transmembrane region of human IFN-κ protein using the TMHMM program, and the results showed that the N-terminal amino acid (aa) at positions 15–37 was a potential transmembrane region, while the N-terminal aa at positions 1–27 was predicted to be IFN-κ signal peptides [Figure 2A]. Based on these results, we hypothesized that aa at positions 28–37 at the N-terminus of the IFN-κ protein was associated with the localization of mature IFN-κ on the cell membrane. To identify the function of the signal peptide and potential transmembrane region, we constructed several mutants of IFN-κ protein with a 6xHis tag, including IFN-κ without 1-27aa (signal peptide), without 1-37aa at the N terminus, or changed signal peptide (replaced with TFF2 signal peptide), or changed signal peptide plus losing 28-37aa [Figure 2B], and detected their expression, location, and function. First, these constructs were transfected into 293T cells for 48 hours, and it was found that the expression of IFN-κ in cells was weak, and losing 1-27 aa or 1-37 aa at the N-terminus further reduced IFN-κ expression, but the replacement of signal peptide enhanced the expression of IFN-κ. In contrast, the expression of IFN-κ and its mutants in the supernatants was hardly detected [Figure 2C].
Next, we examined the localization of these constructs and found that IFN-κ, IFN-κ-sc, and IFN-κ-scΔ28-37 were mainly located on the cell membrane. However, IFN-κΔ1-27 and IFN-κΔ1-37 were not detected on the cell membrane or in the cytoplasm [Figure 2D]. These data indicated that 28-37aa at the N-terminus was not associated with the localization of mature IFN-κ on the cell membrane.
Finally, the ability of these constructs to induce ISGs was evaluated. Only IFN-κ and IFN-κ-sc upregulated the expression of IFITM3, IFN-κΔ1-27, and IFN-κΔ1-37, and IFN-κ-scΔ28-37 lost the ability to induce IFITM3 expression [Figure 2E]. Altogether, this evidence suggests that the signal peptide (1-27aa at the N-terminus) influenced the expression, location, and function of IFN-κ, but 28-37aa at the N-terminus only had an impact on the function of IFN-κ. Therefore, we carried out the sequence alignment of 28-37aa at the N-terminus of human IFN-κ protein with other human IFN-I and found only 30% homology (Supplementary Figure 2, https://links.lww.com/IDI/A10), indicating that this sequence of IFN-κ might not affect the antiviral function of other IFN-Is.
Soluble mature IFN-κ inhibited the replication of respiratory RNA viruses
Does IFN-κ only rely on the cell membrane to exert antiviral effects? We tried to purify IFN-κ protein from supernatants of the IFN-κ -293FT cell line to investigate its function but failed. Therefore, we expressed and purified mature human IFN-κ protein from prokaryotic E. coli to further verify the antiviral effect of soluble IFN-κ. The IFN-κ protein was expressed as an inclusion body. After purification, the purity of the IFN-κ protein was >95% [Figure 3A]. Furthermore, the soluble IFN-κ protein also induced IFITM3 expression in A549 cells at a concentration of at least 100U/mL [Figure 3B]. It also inhibited PR8 and H9N2 virus replication, although the inhibitory effect was not as good as that of IFN-α2 [Figure 3C]. In addition, soluble mature IFN-κ protein was also used to evaluate the induction of ISGs and the effect on SARS-CoV-2 infection in Huh7 cells [Figure 3D], which induced the upregulation of IFITM3 [Figure 3E], and significantly inhibited SARS-CoV-2 replication and release, although the inhibitory effect was also relatively weaker than IFN-α2 [Figure 3F]. These data indicated that soluble mature human IFN-κ also inhibited the replication of respiratory RNA viruses, which was related to the upregulation of ISGs, such as IFITM3.
The expression patterns of ISGs induced by IFN-κ are different from those induced by IFN-α2
To further determine the expression of ISGs induced by IFN-κ and IFN-α2, A549 cells were stimulated with the same concentration of IFN-κ or IFN-α2 for the indicated time (1, 3, 6, 12, or 24 h) and subjected to RNA-Seq analysis [Figure 4A]. The results suggested that the ISGs stimulated by IFN-κ were significantly higher than those stimulated by IFN-α2 after treatment for 1hour, including IFI6, IFI27, BST2, IFITM1, ISG15, OAS2, etc. [Figure 4B]. Quantitative RT-PCR analysis of the expression of several typical ISGs indicated that IFN-κ induced the upregulation of ISGs much faster. However, IFN-κ induced ISG response with lower intensity and shorter half-life than the response of IFN-α2 [Figure 4C]. In addition, we also detected the expression of IFITM3 and MX1 in A549 cells induced by the same concentration of IFN-κ or IFN-α2 by western blotting and found that IFN-α2 induced stronger and more sustained expression of IFITM3 and MX1 than IFN-κ [Figure 4D]. These results suggested that IFN-κ quickly induced upregulation of ISGs, although the effect was relatively weak and transient, while IFN-α2 started later but exerted a stronger and more sustained ISG response.
In conclusion, in contrast to classical IFN-α2 and IFN-β, human IFN-κ exists in two forms, one located on the cell membrane and the other secreted outside the cell. Human IFN-κ preferentially inhibits respiratory RNA virus replication in a membrane-anchored manner. We detected IFN-κ expression in the supernatants of 293FT cells stably expressing IFN-κ. However, its capacity to induce ISG was hardly detected. There may be two reasons for this phenomenon: one is that the IFN-κ protein is unstable in supernatants due to the broken disulfide bonds between cysteine and cysteine, and the other is that the IFN-κ protein is modified by glycosylation. The NetNGlyc-1.0 Server predicted at least six potential N-glycosylation sites. Moreover, the molecular weight of IFN-κ in supernatants was larger than that in cells [Figure 1A, right]. Both may affect the antiviral function of IFN-κ. Notably, if mature IFN-κ is expressed in prokaryotic systems, it also inhibits respiratory RNA virus replication. Compared with classical IFN-I (such as IFN-α2), human IFN-κ induces an effect more rapidly, but its intensity is weaker and its half-life is shorter.
Our results align with those of a previous study in which IFN-κ was detected on the cell surface and exhibited cell-associated antiviral activity. We further validated that the lack of a signal peptide (1-27aa at the N-terminus) affected the expression, localization, and function of IFN-κ, but signal peptide replacement reversed this result. In addition, 28-37 aa at the N-terminus did not affect the expression and location of IFN-κ. However, the loss of 28-37 aa at the N-terminus disabled its function of stimulating IFITM3. The predicted three-dimensional structure of IFN-κ protein showed that 28-37aa at the N-terminus was located in a α-helix, while 146-158aa (a unique region of IFN-κ different from other type I interferons) was located in the loop region of the C/D α-helices (Supplementary Figure 3, https://links.lww.com/IDI/A11). At present, the function of the unique loop region (146-158aa) is still unknown, and we found that loss of this unique loop region did not affect the expression and function of IFN-κ (Supplementary Figure 4, https://links.lww.com/IDI/A12). Since IFN-κ contains more basic amino acids (35 basic aa vs. 26 acid aa) and its isoelectric point reaches 9.1, we speculated that IFN-κ was located on the cell membrane because these more basic amino acids interact with phosphate groups on the surface of the cell membrane through electrostatic interactions.
Gene location and structure analyses have shown that IFN-κ has evolved separately from other IFN-Is. We hypothesize that IFN-κ or its ortholog evolved earlier than other IFN-Is and may be more ancient IFN-Is; they may exist in single-cell organisms and protect themselves from viral infection without help from other cells. In humans, various subtypes of IFN-I exist. However, the functional differences between the different subtypes remain elusive. This study found that IFN-κ induced the upregulation of ISGs much faster, but with lower intensity and shorter half-life than IFN-α2. We speculate that the IFN-I response occurs in a cascade manner similar to the host immune response. First, IFN-κ located on the cell membrane at mucosal surfaces constitutes the first line of defense for an IFN-I response. When a virus invades, IFN-κ swiftly exerts an antiviral effect on the local mucosal site. If viral replication is not controlled or the virus spreads to the circulation system, the second wave of IFN-I response (such as IFN-α/β) is induced to resist virus invasion, with higher intensity and a longer half-life.
At present, infectious diseases caused by respiratory RNA viruses, including influenza and coronaviruses, frequently occur. IFN-κ inhibits respiratory RNA virus replication and may be a potential broad-spectrum antiviral drug. Compared with IFN-α2, IFN-κ exerted an antiviral effect on the local mucosa in an auto-stimulation or para-stimulation manner, possibly reducing the occurrence of side effects. Therefore, the use of IFN-κ to treat patients with SARS-CoV-2 has been included in the “Comprehensive treatment and management of corona virus disease 2019: expert consensus statement from Shanghai City”.
However, there are still many scientific issues that need to be resolved, such as why IFN-κ in supernatants of eukaryotic cells lose their antiviral function, how IFN-κ is located on the cell membrane, and what is the specific function of the unique loop region. In addition, the distribution of IFN-κ in the respiratory tract and its antiviral effect in vivo need to be validated. Clarification of these questions will help to further understand the origin and evolution of IFN-κ, even IFN-I.
The authors thank prof. Zejun Li at Veterinary Research Institute from Chinese Academy of Agricultural Sciences for providing H9N2 virus generously, and the group of prof. Lin Li at the Institute of Microbial Epidemiology from the Academy of Military Medical Sciences for carrying out the SARS-CoV-2 infection assay at biosafety level III laboratory.
This work was supported by the National Science Foundation of China (No. 82071788), and Three-year Action Plan for Shenkang Clinical Research from Shanghai Municipal Health Commission (SHDC2020CR3011A), and the Special Medical Innovation Research Project of “Scientific and Technological Innovation Action Plan” of Shanghai Municipal Commission of Science and Technology in 2020 (20Y11900500).
Jianqing Xu configured this study. Jianqing Xu, Xiaoyan Zhang, and Shuye Zhang designed the study. Jianqing Xu and Xiaoyan Zhang had full access to all data in the study, and took responsibility for the data's integrity and the accuracy of the data analysis. Weihui Fu carried out most of the assay, analyzed data, and wrote the manuscript. Jianqing Xu contributed to the critical revision of the manuscript. Peng Sun analyzed the RNA-Seq data and validated ISGs’ expression. Jun Fan constructed IFN-κ stably expressing cells. Longfei Ding, Songhua Yuan, and Miaomiao Zhang amplified and titrated the influenza virus. Guanxing Zhai and Chenli Qiu provided help in the construction of cloning or Flow Cytometry. All authors reviewed and approved the final version.
Conflicts of Interest
Editor note: Jianqing Xu and Shuye Zhang are the Editors of Infectious Diseases & Immunity. The article was subject to the journal's standard procedures, with peer review handled independently of those members and their research groups.
. LaFleur DW, Nardelli B, Tsareva T, et al. Interferon-kappa, a novel type I interferon expressed in human keratinocytes. J Biol Chem 2001;276(43):39765–39771. doi: 10.1074/jbc.M102502200.
. Habiger C, Jäger G, Walter M, et al. Interferon kappa inhibits human papillomavirus 31 transcription by inducing Sp100 proteins. J Virol 2016;90(2):694–704. doi: 10.1128/JVI.02137-15.
. Li Y, Song Y, Zhu L, et al. Interferon kappa is important for keratinocyte host defense against Herpes Simplex Virus-1. J Immunol Res 2020;2020:5084682. doi: 10.1155/2020/5084682.
. Santhakumar D, Iqbal M, Nair V, et al. Chicken IFN kappa: a novel cytokine with antiviral activities. Sci Rep 2017;7(1):2719. doi: 10.1038/s41598-017-02951-2.
. Guo Y, An D, Liu Y, et al. Characterization and signaling pathway analysis of interferon-kappa in bovine. Dev Comp Immunol 2017;67:213–220. doi: 10.1016/j.dci.2016.09.018.
. Gao M, Guo Y, Du J, et al. Evolutional conservation of molecular structure and antiviral function of a type I interferon, IFN-kappa, in poultry. Dev Comp Immunol 2018;89:44–53. doi: 10.1016/j dci.2018.08.004.
. Fu W, Liu Y, Xia L, et al. A clinical pilot study on the safety and efficacy of aerosol inhalation treatment of IFN-kappa plus TFF2 in patients with moderate COVID-19. EClinicalMedicine 2020;25:100478. doi: 10.1016/j.eclinm.2020.100478.
. Fu W, Liu Y, Liu L, et al. An open-label, randomized trial of the combination of IFN-kappa plus TFF2 with standard care in the treatment of patients with moderate COVID-19. EClinicalMedicine 2020;27:100547. doi: 10.1016/j.eclinm.2020.100547.
. Harris BD, Schreiter J, Chevrier M, et al. Human interferon-ε and interferon-κ exhibit low potency and low affinity for cell-surface IFNAR and the poxvirus antagonist B18R. J Biol Chem 2018;293(41):16057–16068. doi: 10.1074/jbc.RA118.003617.
. He Y, Fu W, Cao K, et al. IFN-kappa suppresses the replication of influenza A viruses through the IFNAR-MAPK-Fos-CHD6 axis. Sci Signal 2020;13(626):eaaz3381. doi: 10.1126/scisignal.aaz3381.
. Buontempo PJ, Jubin RG, Buontempo CA, et al. Antiviral activity of transiently expressed IFN-kappa is cell-associated. J Interferon Cytokine Res 2006;26(1):40–52. doi: 10.1089/jir.2006.26.40.
. Zhang H, Duan ZJ, Zhu JG, et al. Cloning, expression and purification of interferon-kappa, a novel human interferon, and its antiviral activity. Zhonghua Shi YanHe Lin Chuang Bing Du Xue Za Zhi 2005;19(3):223–226. doi:10.3760/cma.j.issn.1003-9279.2005.03.006.
. Shanghai Clinical Treatment expert Group for Corona Virus Disease 2019. Comprehensive treatment and management of corona virus disease 2019: expert consensus statement from Shanghai City. Chin J Infect Dis 2020;38(3):134–138. doi:10.3760/cma.j.issn.1000-6680.2020.03.002.