Genome-wide CRISPR screen identifies STK11 as a critical regulator of sialic acid clusters important for influenza A virus attachment

Graphical abstract

STK11 facilitates influenza A virus (IAV) attachment by activating the RhoA signaling pathway to promote F-actin stress fiber formation, which maintains sialic acid clusters in a disordered, interdigitated, and irregular spatial organization favorable for multivalent HA binding. STK11 deficiency suppresses RhoA-mediated F-actin stress fiber formation, resulting in ordered receptor clustering and reduced viral attachment efficiency.

Highlights

• •A genome-wide CRISPR knockout screen identifies host genes critical for SIV replication.
• •STK11 deficency effectively restricts IAV replication in vitro and in vivo.
• •STK11 is essential for IAV attachment.
• •STK11 modulates actin stress fibers formation to maintain the spatial organization of sialic acid clusters.

Abstract

Introduction

Swine influenza virus (SIV) is a highly contagious respiratory pathogen in pigs that causes substantial economic losses in the pig industry. Importantly, pigs act as “mixing vessels” for diverse influenza A viruses (IAVs), facilitating the emergence of novel pandemic strains through reassortment, which represents a continuous global public health threat. IAV replication relies heavily on host cellular machinery, underscoring the importance of elucidating virus-host protein interactions for the development of targeted antiviral therapeutics.

Objectives

This study aims to identify host genes required for SIV replication via a genome-wide CRISPR screen and elucidate the mechanism by which STK11 modulates viral replication.

Methods

A pig genome-scale CRISPR knockout (PigGeCKO) screen was performed in newborn pig trachea (NPTr) cells to identify host genes required for SIV replication. Candidate genes were further validated by generating knockout cell lines using CRISPR/Cas9-mediated gene editing, followed by assessing their impact on IAV replication. The specific lifecycle stage regulated by STK11 and its mechanistic role in viral attachment were determined via Western blotting, confocal microscopy, transmission electron microscopy, and stimulated emission depletion (STED) imaging. In vivo validation of STK11 knockdown effects on IAV replication was conducted in BALB/c mice treated with STK11-targeting siRNA, with outcomes evaluated by survival analysis, body weight monitoring, lung viral titers quantification, immunofluorescence, and histopathology.

Results

STK11 promotes replication of different IAV subtypes in vitro, and STK11 knockdown significantly suppresses SIV replication in vivo. Mechanistically, STK11 depletion impairs viral attachment by altering the organization of sialic acid clusters, mediated through reduced intracellular actin stress fibers via inhibition of RhoA signaling pathway.

Conclusion

We identify STK11 as a novel regulator of IAV attachment and elucidate its mechanistic role in facilitating viral entry. These findings highlight the potential of STK11 to serve as an ideal antiviral target against IAV infection.

Introduction

Swine influenza virus (SIV), a member of influenza A virus (IAV), is responsible for causing highly infectious respiratory disease in pigs, resulting in substantial economic losses to the pig industry worldwide [[1], [2], [3]]. Due to the presence of both avian (α2,3-linked) and human (α2,6-linked) sialic acid receptors in their respiratory tracts, pigs are regarded as a pivotal “mixing vessel” for various IAVs, providing a conducive platform for the emergence of new strains through gene reassortment and variation. In addition, SIV, as well as human-adapted influenza viruses that circulate stably in human populations, are capable of bidirectional transmission between pigs and humans. This sustained cross-species transmission within the pig reservoir accelerates viral evolution through reassortment events, thereby significantly increasing the risk of generating novel variants with enhanced zoonotic potential and pandemic threat [[4], [5], [6]]. For instance, the 2009 swine-origin H1N1 pandemic led to 123,000 to 203,000 fatalities worldwide due to respiratory complications [7]. Since then, novel reassortant strains derived from H1N1pdm09 and endemic SIV lineages have been identified in different countries [8,9]. Furthermore, incidents of novel SIV strains overcoming the interspecies barrier to cause human infections have also been documented in recent years, raising serious concerns about the potential emergence of future pandemics [[10], [11], [12], [13]]. These observations highlight the great importance of investigating SIV-host interactions in understanding viral ecology and developing effective countermeasures.

To mitigate the threat of influenza pandemics, efforts to develop antiviral strategies have been ongoing since the 1980s. Currently, several antiviral drugs have been approved by the FDA, such as neuraminidase inhibitors (e.g., oseltamivir) and RNA polymerase inhibitors (e.g., baloxavir marboxil), both of which target critical stages of the viral replication cycle [14,15]. However, the high mutation rate and frequent antigenic shifts continually drive the emergence of drug-resistant variants and immune escape, underscoring the urgent need for innovative therapeutic approaches. Host-targeted antivirals have emerged as a promising alternative, given the strong dependence of IAV replication on host cellular pathways [16,17]. The viral entry process is a critical step in the infectious cycle. IAV attachment mediated by haemagglutinin (HA) and sialic acid receptors is the crucial step to initiate the whole process [18], and determines the host susceptibility, transmission, and pathogenicity of the virus [19,20]. Several host factors have been identified that modulate viral attachment and invasion. For instance, C1QTNF5 and CEACAM6 facilitate viral attachment through interactions with HA or neuraminidase (NA) [21,22]. Sialic acids, typically located at the terminal positions of glycoproteins and glycolipids, serve as primary receptors for influenza virus and many other pathogens [23,24]. Host proteins such as SLC35A1, B3GAT1, and APOE modulate the expression of sialic acid receptors, thereby influencing viral attachment [[25], [26], [27]]. Additionally, B4GALNT2 inhibits viral attachment by introducing steric hindrance that interferes with receptor availability [28]. These insights have led to the development of antiviral strategies aimed at blocking the viral invasion, such as sialic acid-mimicking peptides or polymers that competitively block HA binding, and sialidase mimetics that remove sialic acid residues from decoy receptors [29,30]. Despite these advances, the molecular mechanisms underlying influenza virus invasion remain to be fully characterized.

Serine/threonine kinase 11 (STK11), also known as liver kinase B1 (LKB1), was initially characterized as a tumor suppressor gene whose inactivating mutations are implicated in the pathogenesis of Peutz–Jeghers syndrome [31]. This multifunctional kinase is ubiquitously expressed across various tissues and plays a pivotal role in various cellular processes, including cellular metabolism, energy homeostasis, and cell polarity. It exerts these functions primarily through the phosphorylation and activation of AMP-activated protein kinase (AMPK) family members [32]. Moreover, STK11 also plays crucial roles in viral infections, such as hepatitis B virus (HBV), human T-cell leukemia virus type 1 (HTLV-1), and dengue virus (DENV) [[33], [34], [35]]. Nevertheless, the role of STK11 in virology remains poorly characterized, with no existing evidence supporting its involvement in the early viral entry process or influenza virus replication.

Considering the significant threat posed by the SIV to both the swine industry and global public health, and its strict dependence on host cellular machinery for efficient replication, the identification of critical host factors required for viral replication is essential for the development of host-directed antiviral strategies. This study aims to systematically identify host genes required for SIV replication through a pig genome-scale CRISPR knockout (GeCKO) screen in newborn pig trachea (NPTr) cells. Among the candidates, STK11—a gene not previously linked to influenza virus biology—was selected for further investigation. We validated its role in SIV replication, mapped its involvement across viral life cycle stages, and explored the underlying molecular mechanisms. These findings offer new insights into host–IAV interactions and position STK11 as a promising target for host-directed antiviral strategies.

Materials and methods

Cells and viruses

Porcine kidney epithelial cells (PK-15) and Madin-Darby canine kidney (MDCK) cells preserved in the laboratory, along with the newborn pig trachea (NPTr) cells from Prof. Wendy Barclay (Imperial College in London), were maintained in Dulbecco’s modified Eagle’s medium (DMEM). The human embryonic kidney 293 T (HEK293T) cells from ATCC (Manassas, VA, USA) were maintained in RPMI 1640 medium. Both DMEM and RPMI 1640 were from HyClone (Beijing, China). To all the above culture media, 10 % fetal bovine serum (FBS) from Excell (Wuhan, China) was added, and the cells were cultured at 37 °C with 5 % CO₂ in a humidified environment.

The study utilized the following viruses: A/Swine/Hubei/221/2016 (HuB, H1N1), A/Hunan/42443/2015 (HuN, H1N1), A/Swine/Henan/F26/2017 (F26, H1N1), A/Puerto Rico/8-SV14/1934 (PR8, H1N1), A/Chicken/Shanghai/SC197/2013 (SH13, H9N2), and A/Chicken/Hubei/115/2016 (115, H9N2). The GFP-expressing Newcastle disease virus (NDV-GFP) was generously gifted by Associate Prof. Xueying Hu (Huazhong Agricultural University). Propagation of the viruses was carried out in 10-day-old embryonated specific pathogen-free (SPF) chicken eggs from Boehringer Ingelheim Vital Biotechnology Co. Ltd (Beijing, China), followed by 50 % tissue culture infectious dose (TCID₅₀)-based titration in MDCK cells.

Antibodies and reagents

The study utilized the following antibodies and reagents: Anti-STK11/LKB1 rabbit polyclonal antibody (pAb), anti-GAPDH and RhoA mouse monoclonal antibody (mAb) (Cat No. 10746–1-AP, No. 60004–1-Ig and 66733–1-Ig, Proteintech, Wuhan, China); anti-HA, M1 and NP rabbit pAbs (Cat No. GTX127357, GTX125928 and GTX125989, GeneTex, USA); anti-Clathrin heavy chain rabbit pAb, Rabbit control IgG, FITC goat anti-mouse IgG and Cy3 goat anti-Rabbit IgG (Cat No. A12423, AC005, AS001, and AS007, ABclonal, Wuhan, China); anti-phospho-MRLC1 (p-MLC) (Ser19) and MRLC1 (MLC) rabbit pAb (Cat No. AF5443 and AF5423, Affinity, Changzhou, China); FITC goat anti-rabbit IgG and Cy3 Goat Anti-mouse IgG (Cat No. 202051, 202048, Biopm, Wuhan, China); anti-HA mouse mAb (Cat No. M180-3, MBL, Beijing, China); Protein A/G Magnetic Beads (Cat No. HY-K0202, MCE, Shanghai, China); DAPI (4′,6-diamidino-2-phenylindole, 1:1,000 dilution) (Cat No. C1002, Beyotime, China); Biotinylated Sambucus Nigra Lectin (SNA) and Maackia Amurensis Lectin I (MAL I) (Cat No. B-1305–2 and B-1315–2, Vector Lab, Burlingame, CA, USA); Alexa Fluor® 647 Streptavidin (Cat No. 405237, Biolegend, CA, USA).

Plasmids

The lentiGuide-Puro and lentiCRISPR v2 vectors,  as well as the lenti-sgRNA-EGFP vector constructed with paired oligonucleotides of pig sgRNA pools, were preserved in our laboratory. The full-length pig STK11 and the Rho-binding domain (RBD) of Rhotekin were amplified from cDNA extracted from the NPTr cells. The pig gene STK11 underwent digestion with the restriction enzymes EcoR I and Kpn I, and then was ligated into the eukaryotic expression vector p3xFlag-CMV (Flag-STK11). The Flag-STK11(m) in which the PAM sequence was synonymously mutated was also constructed. Meanwhile, the pig gene STK11 underwent digestion by EcoR I and Xho I, and subsequently inserted into the eukaryotic expression vector pCAGGS-HA (HA-STK11). The Rhotekin-RBD was cloned into pGEX-KG at EcoR I and Hind III sites (GST-Rhotekin-RBD). STK11(m) coupled with Flag-tag was cloned into pLenti6.4-cmv-halo (kindly provided by Prof. Shuhong Zhao, Huazhong Agricultural University) by homologous recombination (pLenti6.4-cmv-halo-Flag-STK11). The viral HA gene of HuB H1N1 was cloned into the pSP72 vector (kindly provided by Prof. Yinhua Huang, China Agricultural University) at Mlu Ⅰ and Pme Ⅰ sites.

Transfection

Transient transfection was conducted with either Lipofectamine™ 2000 (Cat No. 11668019, Invitrogen, USA) or Lipo8000™ Transfection reagent (Cat No. C0533, Beyotime, Shanghai, China). For Lipofectamine™ 2000, plasmids and reagents were diluted in Opti-MEM, followed by a brief 5-minute incubation, thorough mixing, and a subsequent 20-minute incubation to form complexes of DNA and Lipofectamine™ 2000, which were transferred to the cells. Cell medium was refreshed after 6 h. For Lipo8000™, diluted plasmids were mixed directly with the reagent and added to the cells without additional incubation.

The pig genome-wide sgRNA library design and the sgRNA library plasmid construction

The sgRNA library was designed and constructed as previously described [36]. Briefly, the sgRNA library was designed by CRISPR-offinder. Selected sgRNAs were prioritized for targeting the first 50 % of the open reading frames and the miRNA hairpin region. To ensure specificity, the potential of off-target was also minimized by allowing up to three nucleotides of mismatch in the 20mer targeting region. Furthermore, any overlap between sgRNAs directed at the same target was avoided. The sgRNA library was then synthesized and cloned into the lenti-sgRNA-EGFP vector.

sgRNA library lentivirus packaging and NPTr-GeCKO library Generation

To produce lentivirus, co-transfection of library plasmid, pMD2.G, and psPAX2 was performed in HEK293T cells in 100 mm dishes using Lipofectamine™ 2000 as described above. At 60 h following transfection, supernatants were collected, filtered, and ultracentrifuged (153,700 × g, 4 °C, 2.5 h). The resulting virus pellets were resuspended in phosphate buffered saline (PBS) (Cat No. SH30256.01, HyClone, Shanghai, China) and stored at −80 °C.

To generate the NPTr-GeCKO library, approximately 2 × 10^⁸ NPTr-Cas9 cells plated in 100 mm dishes were exposed to the library lentiviruses at a multiplicity of infection (MOI) of 0.3, with a control group of uninfected cells. At 72 hpi, EGFP-positive cells were sorted using flow cytometry and then plated into 100 mm dishes to establish NPTr-GeCKO cells. Genomic DNA extraction from ∼ 10^⁷ NPTr-GeCKO cells was performed with the Blood & Cell Culture DNA Midi Kit (Cat. No. 13343, QIAGEN, USA) to amplify sgRNA-containing fragments for assessment of coverage of NPTr-GeCKO library cells.

Genome-wide NPTr-GeCKO screening

For the CRISPR screening, approximately 1.0 × 10^⁸ NPTr-GeCKO and NPTr-Cas9 cells were simultaneously challenged with HuB H1N1 (MOI = 0.01). At 72 hpi, following the complete death of NPTr-Cas9 cells, the surviving cells were replated for further expansion and infection. This process was repeated four more times, with the last two rounds of screening conducted at an MOI of 0.1. Genomic DNA was isolated from cells refractory to virus infection collected from the second to the fifth screening rounds to amplify sgRNA-containing sequences for deep sequencing, followed by MAGeCK-based data analysis.

Construction of polyclonal knockout cells for candidate genes

Candidate gene-targeting sgRNAs from the sgRNA library were inserted into lentiGuide-Puro and then packaged into lentivirus as described above (specific sgRNA sequences were listed in Table S2). The lentivirus was then used to infect NPTr-Cas9 cells. After 3 days of infection, cells were subjected to 1.5 μg/mL puromycin selection pressure over 7 days, resulting in the establishment of polyclonal knockout cell lines. The genomic DNA of polyclonal knockout cells was isolated and served as a template to amplify genome fragments containing the sgRNA targeting site for sequencing analysis.

Generation of STK11 knockout cells

To generate STK11-KO cells in NPTr, the corresponding polyclonal knockout cells mentioned above were subjected to limiting dilution in 96-well plates for monoclonal isolation. To establish STK11-KO cells in PK-15, STK11 sgRNA was inserted into the lentiCRISPRv2 plasmid, followed by lentivirus production as described above. Next, PK-15 cells were subjected to lentivirus infection for 3 days, after which cells underwent puromycin-based selection at 1.5 μg/ml, and monoclonal isolation was conducted. Ultimately, the knockout efficiencies for both cell lines were confirmed through sequencing and Western blot assay.

Generation of STK11 revertant cells

To generate STK11-KO-STK11 revertant cells, the pLenti6.4-cmv-halo-Flag-STK11 plasmid was constructed. Then, co-transfection of this construction with pMD2.G and psPAX2 was carried out in HEK293T cells to produce lentivirus. Next, STK11-KO cells were exposed to the lentivirus for 3 days, followed by 7 days of 15 μg/mL Blasticidin S selection to obtain polyclonal revertant populations. Subsequently, the monoclonal cells were acquired as previously mentioned and validated via Western blot analysis.

Cell viability determination

Cells were cultivated in 96-well plates to assess their viability. At specified times, cells in each well received 10 μL of CCK-8 reagent, underwent incubation at 37 °C for 1 h in the dark, and absorbance at 450 nm was detected.

Virus infection and titration

NPTr or PK-15 cell lines in 12-well plates were infected with IAV at 37 °C for 1 h. Then, PBS washing and incubation in 0.1 μg/mL tosylsulfonyl phenylalanyl chloromethyl ketone (TPCK)-supplemented DMEM at 37 °C were performed. Supernatants were collected at specific times, diluted in DMEM, and used for MDCK cell infection in 96-well plates under 37 °C conditions for 1 h. Cells were sustained in DMEM with 0.25 μg/mL of TPCK after the inoculum was removed and PBS washes were performed. At 72 hpi, virus titers were assessed by TCID₅₀, determined through the Reed-Muench method.

Ethics statement

All experiments involving animals were conducted according to the ethical policies and procedures approved by the Hubei Administrative Committee for Laboratory Animals (Approval No. HZAUMO-2023–0317). Mice were housed in individually ventilated cages under controlled environmental conditions, with access to environmental enrichment to promote well-being. All invasive procedures were performed under appropriate anesthesia to minimize pain and distress. Humane endpoints were strictly observed: animals exhibiting a body weight loss exceeding 30 % were euthanized using carbon dioxide inhalation, followed by confirmation of death, in compliance with institutional and national standards for animal welfare.

Animal experiments in vivo

Female SPF BALB/c mice, aged 6 weeks, were nasally administered PBS, Control siRNA (siControl), and siRNA targeting STK11 (siSTK11) (siRNAs were synthesized by GenePharma with cholesterol conjugation and 2′-O-methyl (2′-OME) modifications) on days −1 and 1, respectively. On day 0, siControl and siSTK11 groups were exposed to 30 PFU of HuN H1N1, while the PBS group received a mock treatment. Daily weight variations and survival were tracked for 14 days (n = 10), with euthanasia performed upon exceeding 30 % weight loss per animal ethics guidelines. On days 3 and 5 following the challenge, lung tissues (n = 3/group) were either fixed in 4 % paraformaldehyde for histopathology/immunofluorescence or homogenized and centrifuged (12,000 × g, 15 min, 4 °C) to collect supernatants for silencing efficacy detection and viral load quantification.

Co-immunoprecipitation and western blot assay

Co-immunoprecipitation (Co-IP) analysis was performed by lysing cells in NP-40 buffer, with a portion of lysates reserved as input controls. The remaining lysates underwent overnight incubation with target-specific antibodies at 4 °C and 2-hour Protein A/G magnetic bead precipitation under rotation. Then, samples were prepared by combining immunoprecipitated complexes and input controls with SDS loading buffer, underwent SDS-PAGE, transferred to nitrocellulose membranes, and blocked prior to sequential incubation with primary/secondary antibodies. Finally, detection of protein-specific immunoblot signals was accomplished with a chemiluminescence imaging system (Tanon-5200, Tanon, Shanghai).

Acid bypass assay

Following seeding in 12-well plates, STK11-KO and control cells were subjected to HuB H1N1 infection (MOI = 0.1) at 4°C for 1 h, followed by two cold PBS washes. Then the cells were exposed to PBS buffers (pH 5.0 or 7.4) for 10 min, followed by incubation in NH₄Cl-supplemented medium (20 mM). After incubating for 8 h at 37 °C, the cells were prepared for analysis via Western blot.

Indirect immunofluorescence analysis

Cells cultivated on glass-bottom culture dishes were infected with IAV for specified time, and then treated with 4 % paraformaldehyde (PFA) for 10 min. Subsequently, permeabilization was accomplished with 0.2 % Triton X-100 for 10 min, and blocking was carried out using 1 % bovine serum albumin (BSA) for 1 h. This was followed by sequential incubation with target-specific antibodies for 2 h, and then with fluorescent-labeled secondary antibodies for 1 h. Nuclear staining was conducted using DAPI for 10 min, after which confocal imaging was performed through confocal microscopy (Nikon STORM, Japan).

Internalization assay

WT and STK11-KO PK-15 cells were subjected to IAV infection (MOI = 10) at 4 °C for 1 h, followed by a 15-minute incubation at 37 °C. Following this, the cells underwent fixation with 4 % PFA for 10 min, permeabilization for 10 min, and blocking for 1 h. Sequential incubations with primary antibodies targeting CLTA and viral proteins (HA/NP) for 2 h, fluorescent-labeled secondary antibodies for 1 h, and DAPI for 10 min preceded confocal microscopy analysis (Nikon STORM system).

Virus or lectin binding assay

For the virus binding assay, WT and STK11-KO PK-15 cells underwent IAV (MOI = 100) or NDV-GFP (MOI = 10) infection on ice for 1 h to allow attachment without internalization. For IAV-infected cells, post-fixation with 4 % PFA, sequential incubations with HA or NP-specific antibodies for 2 h, and fluorescent-labeled secondary antibody for 1 h preceded nuclear staining with DAPI for 10 min. NDV-GFP-infected cells underwent staining with DAPI directly post-fixation with 4 % PFA. Virus binding was assessed through confocal microscopy (STORM, Nikon, Japan) or flow cytometry (Cytoflex LX, Beckman Coulter, CA, USA).

The lectin binding assay involved cell fixation with 4 % PFA for 10 min, incubation with SNA or MAL Ⅰ (20 µg/mL) for 1 h on ice, and subsequently streptavidin-Alexa Fluor® 647 (1 µg/mL) for 1 h. Cells were DAPI-stained for 10 min, and binding of lectin was observed using confocal microscopy (STORM, Nikon, Japan) and STED super-resolution microscopy (Stedycon, Abberior, Gottingen, Germany).

Transmission electron microscope

Following plating in 6-well plates, WT and STK11-KO PK-15 cells were subjected to HuB H1N1 infection (MOI = 100) via incubation on ice. Following two washes, the cells were fixed using 2.5 % glutaraldehyde, and then examined using a transmission electron microscope (FEI, Tecnai G2 20 TWIN, 200 kv).

Rhotekin-RBD pulldown assay

Following expression in Escherichia coli and purification using glutathione-sepharose 4B beads (Solarbio, Beijing, China), part of the bead-coupled recombinant GST and GST-Rhotekin-RBD proteins were verified by Coomassie blue staining. The remaining purified bead-coupled proteins were rotated with cell lysates from WT, STK11-KO, and STK11-KO-STK11 cells prepared in NP-40 lysis buffer with rotation for 2 h at 4 °C. Following the incubation period, beads underwent five washes, and captured proteins underwent western blotting.

Statistical analysis

GraphPad Prism 8.0 was applied for statistical and graphical analyses, with results presented as mean ± standard deviation (SD) from independent biological replicates. An unpaired two-tailed Student's t-test determined statistical significance (ns, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001, ****; p < 0.0001).

Results

CRISPR-based genome-wide knockout screening reveals host genes crucial for SIV replication in newborn pig trachea cells

To reveal host genes crucial for SIV replication, a genome-wide knockout screening based on cell survival was performed in newborn pig trachea (NPTr) cells transduced with the PigGeCKO sgRNA library using A/Swine/HuBei/221/2016 (HuB, H1N1) virus, as previously established [36]. The highly stringent screens for HuB virus were performed by respectively exposing NPTr-Cas9 and PigGeCKO cells to viral infection until nearly all NPTr-Cas9 cells had died. Virus-resistant cell populations were collected and expanded for an additional four consecutive rounds of challenge with HuB virus. The remaining virus-resistant cells of each round were expanded for genomic DNA extraction, sgRNA library sequencing, and data analysis (Fig. 1A).

Fig. 1.

CRISPR-based genome-wide knockout screening reveals host genes crucial for SIV replication in newborn pig trachea cells. (A) Flowchart of the porcine genome-wide CRISPR screen performed in NPTr cells to discover host genes required for SIV replication. (B) Scatter plot showing enrichment of sgRNA-targeted genes across the 2nd to 5th rounds of screening. (C) Venn diagram depicting the overlap among the top 45 genes identified from the 2nd to 5th rounds of screening. (D-G) Effects of candidate gene knockouts on the replication of different IAV strains. The polyclonal knockout cells were respectively infected with (D) HuB H1N1 (MOI = 0.01), (E) F26 H1N1 (MOI = 0.01), (F) PR8 H1N1 (MOI = 0.01), or (G) SH13 H9N2 (MOI = 0.01), and the viral titers were determined by TCID₅₀ assay at 24 hpi (n = 3). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (ns, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001).

Top candidates from the CRISPR screening were prioritized through the model-based analysis of genome-wide CRISPR/Cas9 knockout (MAGeCK) read counts (Fig. 1B and Table S1). To mitigate the potential loss of key candidates due to inherent technical variations between screens, we adopted a cross-round comparison strategy, prioritizing genes that appeared in multiple rounds. This analysis identified 39 overlapping candidates among the top 45 sgRNA targets across the last four rounds of screening results (Fig. 1C and Table S1). Notably, the well-established host factors implicated in IAV replication, namely the CMP-sialic acid transporter SLC35A1, the V-type ATPase assembly regulator WDR7, ATP6AP2, and ATP6V1F, were among the top-ranked overlapping candidates [25,37,38], thereby confirming the validity of this screen. Next, 12 candidates were selected from the top 39 overlapping genes for further validation. SLC35A1, a well-established regulator for IAV replication that facilitates viral attachment by mediating the biosynthesis of both α-2,6- and α-2,3-linked sialic acid receptors, was included as a positive control [25]. The results showed that HuB virus replication was significantly inhibited in polyclonal knockout cells of these candidates constructed by CRISPR/Cas9 technology (Fig. 1D and Fig. S1A, B). We also found that the replication of other IAV strains, including A/Swine/Henan/F26/2017 (F26, H1N1), A/Puerto Rico/8/1934 (PR8, H1N1), and A/chicken/Shanghai/SC197/2013 (SH13, H9N2), was also markedly inhibited in most polyclonal knockout cells (Fig. 1E-G), verifying the conserved functions of these candidate genes across different IAV strains.

STK11 promotes IAV replication

As STK11 and SLC35A1 polyclonal knockout cells exhibited the most pronounced and comparable inhibitory effect on all tested IAV strains, we were interested in the underlying mechanism by which STK11 regulates SIV replication, as its role in IAV infection has not been previously reported. STK11 monoclonal knockout (STK11-KO) cells and control cells expressing a non-targeting control sgRNA were generated in NPTr cells, which were confirmed via protein analysis (Fig. 2A). Moreover, the cell viability remained unaffected by STK11 deficiency (Fig. 2B). Following HuB virus infection, the STK11-KO cells exhibited comparable inhibitory effects on SIV replication as SLC35A1-KO cells (Fig. 2C), confirming the critical role of STK11 in SIV replication. In addition, STK11-KO cells also suppressed the replication of other IAV strains, including swine-origin (F26, H1N1), human-origin (PR8 H1N1), and avian-origin (SH13 H9N2) influenza virus (Fig. 2D-F). To further confirm the promoting effect of STK11 on SIV replication, control or STK11-KO cells complemented with exogenous STK11 were infected with HuB virus. The results suggested that complementation of exogenous STK11 significantly promoted or restored HuB virus replication in control and STK11-KO cells, respectively (Fig. 2G), suggesting that STK11 promoted SIV replication. STK11 expression and localization were evaluated at various time points post-IAV infection, revealing no significant alterations (Fig. S2A, B).

Fig. 2.

STK11 promotes SIV replication. (A) Western blot analysis of STK11 protein expression in control and STK11-KO NPTr cells. (B) Cell viability of control and STK11-KO NPTr cells assessed by CCK-8 assay (n = 8). (C-F) Viral replication kinetics in control and STK11-KO NPTr cells. Control and STK11-KO NPTr cells were infected with (C) HuB H1N1, (D) F26 H1N1, (E) PR8 H1N1, or (F) SH13 H9N2, and viral titers in supernatants were measured at 12, 24, and 36 hpi via TCID₅₀ assay (n = 3). (G) Overexpression or complementation of exogenous STK11 promotes SIV replication. Control and STK11-KO NPTr cells were transfected with Flag-STK11, a synonymous PAM-mutated STK11 construct (Flag-STK11(m)), or an empty Flag vector, followed by infection with HuB H1N1 (MOI = 0.01). Viral titers were determined by TCID₅₀ assay, and protein expression was assessed by western blotting at 24 hpi (n = 3). (H) Western blot detection of STK11 protein expression in WT and STK11-KO PK-15 cells. (I) Cell viability determination of WT and STK11-KO PK-15 cells by CCK-8 assay (n = 8). (J) Replication of HuB H1N1 virus in WT and STK11-KO PK-15 cells. WT and STK11-KO PK-15 cells were infected with HuB H1N1 (MOI = 0.01), and viral titers in supernatants were quantified via TCID₅₀ assay at 12, 24, and 36 hpi (n = 3). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (ns, not significant; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001).

To validate the role of STK11 beyond respiratory cells, STK11-KO porcine kidney epithelial (PK-15) cells were generated without compromising cell viability (Fig. 2H, I). Consistently, STK11-KO cells exhibited a significant reduction in HuB viral titers compared to WT cells (Fig. 2J), reinforcing STK11 as a critical host factor for SIV replication. To assess species specificity, a polyclonal STK11 knockdown (KD) cell line was established in human embryonic kidney (HEK) 293 T cells without affecting the cell viability (Fig. S3A, B). Upon PR8 virus infection, STK11-KD 293 T cells showed similarly reduced IAV replication relative to WT cells (Fig. S3C), indicating that STK11′s proviral function is conserved across species.

STK11 silencing suppressed SIV replication in vivo

As STK11 deficiency inhibited SIV replication in vitro, we next sought to validate its role during SIV infection in vivo. Female BALB/c mice aged six weeks were inoculated intranasally with cholesterol-coupled and 2′-Ome modified siRNA targeting STK11 (siSTK11) or a negative control siRNA (siControl) at days −1 and 1, followed by intranasal challenge with 30 PFU of human isolate A/Hunan/42443/2015 (HuN, H1N1) at day 0 (Fig. 3A). Efficient STK11 knockdown was confirmed by western blot analysis of lung tissue collected at 3 days post-infection (dpi) (Fig. 3B). Continuous monitoring of disease progressions revealed that STK11-knockdown mice exhibited a reduced weight loss and a 20 % higher survival rate compared to siControl-treated mice (Fig. 3C, D). Moreover, the virus loads in the lungs of STK11-knockdown mice were markedly decreased relative to siControl-treated mice at 3 and 5 dpi (Fig. 3E). The histopathological detection revealed attenuated hemorrhage, edema, and inflammatory infiltration in lungs of STK11-knockdown mice relative to siControl-treated mice (Fig. 3F). Consistently, diminished NP expression was observed in the lung tissues of STK11-knockdown mice (Fig. 3G). In addition, we detected the effect of IAV infection on STK11 expression in vivo. The result revealed no significant alteration in STK11 expression level post-infection (Fig. S4). Overall, these results further proved the critical role of STK11 during SIV infection in vivo.

Fig. 3.

STK11 silencing suppresses SIV replication in vivo. (A) Schematic representation of siRNA delivery and HuN H1N1 challenge in mice. (B) The assessment of STK11 expression in siRNA-targeted mice via western blotting. (C, D) Body weight variations and survival rates of mice treated with the indicated siRNAs. Each treatment group had ten mice (n = 10). (E) Viral titers in lung homogenates from siRNA-treated mice at 3 and 5 dpi, determined by TCID₅₀ assay. Each treatment group had three mice (n = 3). (F) Histopathological examination of lungs from siRNA-treated mice at 3 and 5 dpi, scale bar = 200 μm. (G) Immunofluorescence staining of viral NP in lung sections from siRNA-treated mice at 3 and 5 dpi. NP protein showed a red stain, and the nucleus showed a blue stain. Scale bar = 200 μm. Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (**, p < 0.01; ***, p < 0.001).

STK11 is required for efficient IAV attachment

STK11 modulates multiple biological processes primarily through phosphorylating and activating members of the AMPK family, with AMPK representing its canonical and best-characterized substrate [39]. To investigate whether STK11 promotes IAV replication through AMPK, we examined AMPK phosphorylation in STK11-KO cells and observed significant suppression (Fig. S5A). However, reactivation of AMPK using compound 991, a specific allosteric activator, failed to rescue the impaired viral replication observed in STK11-KO cells (Fig. S5B-D). These findings collectively demonstrated that STK11 promoted IAV replication independently of AMPK signaling.

To determine the specific stage of the SIV lifecycle where STK11 is involved, the expression and distribution of NP protein in HuB virus-infected STK11-KO and control cells were detected at 3, 6, and 9 hpi. STK11-KO cells displayed markedly reduced NP signals relative to controls. In contrast to WT cells, where the viral NP protein is predominantly localized to the nucleus, STK11-KO cells exhibit only minimal nuclear accumulation of NP at 3 hpi (Fig. 4A, B and Fig. S6A, B), indicating a defect at an early stage of infection. Next, the cellular localization and expression of viral M1 were evaluated at earlier consecutive time points from 0.5 to 2.5 hpi. Similarly, a substantial decrease of viral M1 signal was observed in STK11-KO cells relative to control cells as early as 0.5 hpi (Fig. 4C, D and Fig. S6C), further confirming the role of STK11 in the early stage of SIV replication.

Fig. 4.

STK11 is required for efficient IAV attachment. (A, B) Expression and subcellular distribution of viral NP. (A) Control and STK11-KO NPTr cells were infected with HuB H1N1 (MOI = 0.1), and the NP expression and distribution were detected by confocal microscopy at 3, 6, and 9 hpi. Scale bar = 20 μm. (B) Quantification of the average fluorescence intensity per cell in (A) was performed using ImageJ software on 15 fields per sample (n = 15). (C, D) Expression and subcellular distribution of viral M1. (C) Control and STK11-KO NPTr cells were infected with HuB H1N1 (MOI = 10). The M1 expression and distribution were detected by confocal microscopy from 0.5 to 2.5 hpi. Scale bar = 20 μm or 10 μm. (D) Quantification of the average fluorescence intensity per cell in (C) was performed using ImageJ software on 10 fields per sample (n = 10). (E, F) Colocalization of viral HA and CLTC. (E) WT and STK11-KO PK-15 cells were infected with HuB H1N1 (MOI = 10), and the distribution of HA and CLTC was examined by confocal microscopy at 15 min post-infection. Scale bar = 5 μm. (F) The colocalization between viral HA and CLTC in (E) was quantified by calculating Mander’s colocalization coefficients using ImageJ software on 15 fields per sample (n = 15). (G, H) Detection of internalized viral particles. (G) WT and STK11-KO PK-15 cells were infected with HuB H1N1 (MOI = 10) on ice for 1 h, and then shifted to 37 °C for 15 min. The uninternalized virus was removed with 0.25 % Trypsin-EDTA, and then the cells were incubated with anti-HA antibody and analyzed by flow cytometry. (H) The median fluorescence intensity in (G) was quantified (n = 3). (I-L) Determination of viral attachment. WT and STK11-KO PK-15 cells were infected with HuB H1N1 (MOI = 100) on ice for 1 h, and then were incubated with an anti-HA antibody for examination by (I) confocal microscopy, scale bar = 20 μm, and (K) flow cytometry. For the flow cytometry assay, SLC35A1-KO cells were included as a positive control. (J) The average fluorescence intensity per cell in (I) was quantified by ImageJ software on 10 fields per sample (n = 10). (L) The median fluorescence intensity in (K) was analyzed (n = 3). (M) Determination of viral attachment by transmission electron microscopy. WT and STK11-KO PK-15 cells were infected with HuB H1N1 (MOI = 100) on ice for 1 h, and then were fixed and observed by transmission electron microscopy. Scale bar = 2 μm or 1 μm. (N) Western blot analysis of Flag-tagged STK11 in STK11-KO and STK11 revertant cells. (O, P) Back-complementation of STK11 restored viral attachment. (O) WT, STK11-KO, and STK11 revertant PK-15 cells were infected with HuB H1N1 (MOI = 100) for 1 h on ice, and the cells were incubated with anti-HA antibodies for confocal imaging. Scale bar = 20 μm. (P) The average fluorescence intensity per cell in (O) was quantified by ImageJ software on 10 fields per sample (n = 10). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (ns, not significant; ***, p < 0.001; ****, p < 0.0001).

Next, to investigate the impact of STK11 deficiency on the endosomal pathway, an acid bypass assay was conducted, allowing direct fusion of pre-attached viral particles with the plasma membrane to bypass the endosomal pathway [40]. As shown in Fig. S6D, exposure to acidic pH did not reverse the inhibitory effect on SIV replication observed in STK11-KO cells, indicating the role of STK11 at the stage preceding internalization. To corroborate this, the co-localization of viral HA and clathrin heavy chain (CLTC) in HuB virus-infected STK11-KO and WT PK-15 cells was examined. It was observed that STK11-KO cells exhibited a marked decrease in both HA signal and co-localization of HA with CLTC compared to WT cells (Fig. 4E, F). Meanwhile, we also detected the internalized virions by flow cytometry. Consistently, significantly reduced internalized virions in STK11-KO cells were observed compared to those in WT cells (Fig. 4G, H and Fig. S6E). Furthermore, a similar reduction in internalized virions was observed in PR8 H1N1 and 115 H9N2 infected STK11-KO cells compared to WT cells (Fig. S6F, G). Together, these findings suggested that STK11 functioned at the step prior to viral internalization.

Next, the role of STK11 in viral attachment was investigated. STK11-KO and WT cells infected with HuB virus were subjected to analysis of viral HA via confocal microscopy and flow cytometry. The HA signal intensity in STK11-KO cells exhibited a significant reduction compared to WT cells, implying that the absence of STK11 impeded SIV attachment (Fig. 4I-L). Similarly, a decrease in viral HA or NP signals of attached virions was observed in PR8 H1N1 and 115 H9N2 infected STK11-KO cells (Fig. S6H, I). Furthermore, transmission electron microscopy observations revealed that virions were dispersed around the cell membrane of STK11-KO cells, while tightly attached to WT cells (Fig. 4M). These results demonstrated the critical role of STK11 in viral attachment.

To further support these findings, a STK11 revertant cell line was generated by stable reintroduction of Flag-tagged STK11 protein into STK11-KO cells, as confirmed by protein analysis (Fig. 4N). Interestingly, the STK11 revertant cells exhibited comparable HA signal intensity of attached virions to that in WT cells (Fig. 4O, P and Fig. S6J), confirming that the diminished viral attachment in STK11-KO cells was specifically attributed to the lack of STK11. Collectively, these findings confirmed that STK11 is a crucial host factor that facilitates efficient influenza virus attachment.

STK11 deficiency alters the spatial organization of sialic acid clusters

IAV attachment is primarily mediated by viral HA and sialic acid receptors [18]. To explore the mechanism by which STK11 facilitates IAV attachment, we first detected the interaction between viral HA and STK11 to evaluate the potential of STK11 as a cofactor in viral attachment. Co-immunoprecipitation (Co-IP) analysis revealed that STK11 did not co-precipitate with viral HA under both conditions of exogenous transfection and HuB virus infection (Fig. 5A, B), suggesting that STK11 did not facilitate SIV attachment by cooperating with viral HA. Next, the impact of STK11 deficiency on the expression of sialic acid receptors was determined. Notably, no reduction in either α2,3- or α2,6-linked sialic acid receptors was observed in STK11-KO cells compared to WT cells (Fig. 5C, D), demonstrating that the decreased SIV attachment caused by STK11 knockout was not attributed to decreased sialic acid receptor expression.

Fig. 5.

STK11 deficiency alters the spatial organizations of sialic acid clusters. (A, B) Co-IP analysis of the interaction between viral HA and STK11 in (A) HEK293T cells transfected with HA-tagged STK11 and pSP72-HA, or (B) PK-15 cells infected with HuB H1N1 (MOI = 0.01). (C, D) Detection of α2, 3 or α2, 6-linked sialic acid receptor abundance on the cell membrane. WT and STK11-KO PK-15 cells were stained with (C) Maackia Amurensis Lectin I (MAL I) or (D) Sambucus Nigra (SNA), which specifically bind α2,3 and α2,6-linked sialic acid receptors, respectively, and analyzed by confocal microscopy. Scale bar = 20 μm. (E) Observation of α2, 3 and α2, 6-linked sialic acid clusters on the cell membrane. WT and STK11-KO PK-15 cells were stained with MAL I or SNA and analyzed by STED microscopy. Scale bar = 2 μm. (F, G) Effect of STK11 knockout on NDV-GFP attachment. (F) WT and STK11-KO PK-15 cells were infected with NDV-GFP for 1 h on ice, and then were fixed and observed by confocal microscopy. Scale bar = 20 μm. (G) The average fluorescence intensity per cell in (F) was calculated by ImageJ software on 10 fields per sample (n = 10). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (****, p < 0.0001).

Efficient IAV attachment relies on multivalent HA-sialic acid interactions, and sialic acid receptors are organized into clusters at the cell surface, providing a platform for such multivalent interactions [41,42]. We next investigated the impact of STK11 deficiency on sialic acid clusters. Stimulated emission depletion (STED) microscopy imaging revealed that α2, 3 and α2, 6-linked sialic acid clusters exhibited regular organized pattern extending outwardly perpendicular to the cell surface with apparent demarcation between clusters on cell membrane of STK11-KO cells, while exhibited disorganized and irregular distributions, interdigitating with surrounding clusters without clear directionality on WT cell membrane (Fig. 5E). To further evaluate the functional relevance of this altered organization, we assessed the attachment of Newcastle disease virus (NDV), another virus utilizing sialic acids as attachment receptors. Consistently, the attachment of NDV was significantly reduced in STK11-KO cells relative to WT cells (Fig. 5F, G). These observations indicated that STK11 deficiency impaired IAV attachment by altering the spatial organization of sialic acid clusters.

STK11 regulates F-actin stress fiber formation via RhoA signaling to maintain the spatial organization of sialic acid clusters and facilitate viral attachment

The intracellular actin cytoskeleton regulates the morphology, density, and nanoscale organization of membrane-associated protein clusters [43,44]. To determine the possible correlation between alteration in the spatial organization of sialic acid clusters caused by STK11 deficiency and actin cytoskeleton dynamics, we first assessed the impact of STK11 on actin cytoskeleton. The results showed that STK11 deficiency markedly impaired intracellular F-actin stress fiber formation compared to WT cells, while the stress fiber formation in STK11 revertant cells was restored to a level comparable to WT cells (Fig. 6A, B), confirming the regulatory role of STK11 in actin cytoskeleton dynamics. To investigate whether the decreased F-actin stress fibers were responsible for the altered spatial organization of sialic acid clusters in STK11-KO cells, we treated cells with Jasplakinolide, a pharmacological inducer of stress fiber formation [45]. Jasplakinolide treatment efficiently restored F-actin stress fibers in STK11-KO cells without affecting cell viability (Fig. 6C, D and Fig. S7). Subsequent STED imaging revealed that Jasplakinolide treatment restored the disordered and interdigitated spatial organizations of sialic acid clusters in STK11-KO cells, phenocopying the pattern observed in DMSO-treated WT cells. Moreover, Jasplakinolide treatment further increased the degree of disorder and interdigitation of sialic acid clusters in WT cells (Fig. 6E). Functionally, Jasplakinolide treatment significantly restored viral attachment in STK11-KO cells and further enhanced attachment in WT cells (Fig. 6F-H). These findings indicated that STK11 regulates actin dynamics to maintain the spatial organization of sialic acid clusters, ultimately facilitating viral attachment.

Fig. 6.

STK11 regulates F-actin stress fiber formation via RhoA signaling to maintain the spatial organization of sialic acid clusters and facilitate viral attachment. (A) Observation of F-actin stress fibers in WT, STK11-KO, and STK11 revertant PK-15 cells stained with fluorescent green phalloidin using confocal microscopy. Scale bar = 20 μm or 10 μm. (B) The number of F-actin stress fibers per field in (A) was quantified by ImageJ software on 15 fields per sample (n = 15). (C) Observation of F-actin stress fibers in STK11-KO cells treated with DMSO or Jasplakinolide (Jas) for the indicated times via confocal microscopy. Scale bar = 20 μm or 10 μm. (D) The number of F-actin stress fibers per cell in (C) was quantified by ImageJ software on 15 cells (n = 15). (E) Observation of sialic acid clusters by STED microscopy. WT and STK11-KO PK-15 cells treated with DMSO or Jas for 24 h were stained with MAL I or SNA and analyzed by STED microscopy. Scale bar = 2 μm. (F-H) The influence of Jas treatment on SIV attachment. WT and STK11-KO PK-15 cells pretreated with DMSO or Jas for 24 h were infected with HuB H1N1 (MOI = 100) for 1 h on ice. (F) Cells were incubated with HA-specific antibody and fluorescent green phalloidin, and visualized by confocal microscopy. Scale bars = 20 μm or 10 μm. (G) The average fluorescence intensity per cell for viral HA in (F) was quantified by ImageJ software on 10 fields per sample (n = 10). (H) Viral protein expression was assessed by western blotting. The ratio of viral HA, M1, or NP to GAPDH was calculated by ImageJ software. (I) Detection of active GTP-RhoA in WT, STK11-KO, and STK11 revertant PK-15 cells by pull-down assay. The ratio of GTP-RhoA or total RhoA to GAPDH was calculated by the ImageJ software. (J) Western blot analysis of MLC and its phosphorylation in WT, STK11-KO, and STK11 revertant PK-15 cells. The ratio of p-MLC or MLC to GAPDH was calculated by ImageJ software. (K) Western blot analysis of LIMK phosphorylation. The ratio of p-LIMK to GAPDH was calculated by ImageJ software. (L)Western blot analysis of RhoA expression in WT, STK11-KO, and STK11-KO-RhoA-Q63L PK-15 cells. (M, N) Constitutively active RhoA (RhoA-Q63L) rescued F-actin stress fiber formation in STK11-KO cells. (M) WT, STK11-KO, and STK11-KO-RhoA-Q63L PK-15 cells were stained with fluorescent green phalloidin and observed by confocal microscopy. Scale bar = 20 μm or 10 μm. (N) Quantification of F-actin stress fiber in (M) was performed by ImageJ software on 15 fields per sample (n = 15). (O, P) Constitutively active RhoA rescued viral attachment in STK11-KO cells. (O) WT, STK11-KO, and STK11-KO-RhoA-Q63L PK-15 cells were infected with HuB H1N1 (MOI = 100) for 1 h on ice, and cells were treated with anti-HA antibodies for confocal imaging. Scale bar = 20 μm. (P) The average fluorescence intensity per cell in (O) was quantified by ImageJ software on 10 fields per sample (n = 10). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (ns, not significant; *, p < 0.05; ****, p < 0.0001).

The formation of intracellular stress fibers is predominantly regulated by small GTPase RhoA and its downstream effector pathways. Activation of RhoA leads to phosphorylation of myosin light chain (MLC), ultimately promoting actin stress fiber assembly. Additionally, activation of LIM domain kinases (LIMKs) stabilizes actin filaments through phosphorylation of cofilin [46,47]. As expected, both total RhoA and its active form GTP-RhoA were dramatically decreased in STK11-KO cells compared to WT cells, while they were rescued in STK11 revertant cells (Fig. 6I). Correspondingly, the downstream MLC and its phosphorylation level were significantly inhibited in STK11-KO cells and were restored in STK11 revertant cells (Fig. 6J). Similarly, the phosphorylation of LIMK was markedly suppressed in STK11-KO cells and restored in STK11 revertant cells (Fig. 6K). To further demonstrate these findings, a STK11-KO cell line expressing constitutively active RhoA (RhoA-Q63L) was constructed (Fig. 6L). We found that re-expression of RhoA-Q63L restored F-actin stress fiber formation and rescued viral attachment in STK11-KO cells (Fig. 6 M−P). Collectively, these findings suggested that STK11 regulates actin cytoskeleton dynamics via RhoA and its downstream pathways to maintain the spatial organization of sialic acid clusters, thus promoting viral attachment.

Discussion

In this study, we generated an NPTr-GeCKO library to identify host dependency factors supporting SIV replication. Furthermore, we illustrated that the absence of STK11 resulted in reduced SIV replication both in vitro and in vivo. Specifically, beyond mammalian cells, knockdown of avian STK11 also suppressed avian influenza virus replication (data not shown), indicating that its proviral function is evolutionarily conserved across species. Subsequent investigations verified that the absence of STK11 altered the spatial arrangements of α2, 3 and α2, 6-linked sialic acid clusters on cell membrane by inhibiting the formation of intracellular actin stress fibers, thereby disrupting IAV attachment (Fig. 7). Collectively, these findings identify STK11 as a key host factor supporting influenza virus entry and underscore its potential as a promising target for host-directed antiviral interventions.

Fig. 7.

Proposed model of STK11-mediated regulation of IAV attachment via the RhoA signaling pathway. (A) STK11 promotes the formation of intracellular F-actin stress fibers by activating small GTPase RhoA and its downstream effectors, including MLC and LIMK. The resulting actin stress fibers exert mechanical forces on the plasma membrane, maintaining sialic acid clusters in a disordered, interdigitated, and irregular spatial configuration. This spatial organization facilitates multivalent interactions between viral HA and sialic acid receptors, enhancing viral attachment. (B) STK11 deficiency impairs RhoA signaling, leading to suppressed F-actin stress fiber formation. The consequent reduction in intracellular F-actin stress fibers decreases mechanical forces exerted on the plasma membrane, reorganizing sialic acid clusters from a disordered and interdigitated pattern into a more ordered and regular pattern. This reorganization reduces the multivalent binding capacity of HA to sialic acid receptors, thereby impairing efficient viral attachment.

In recent decades, CRISPR/Cas9-based genome-scale screening has been widely used in uncovering host dependency factors for diverse viruses, such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), IAV, and human immunodeficiency virus (HIV) [37,[48], [49], [50]]. To date, numerous host genes have been identified to engage in IAV lifecycle by CRISPR/Cas9 screens [26,37,[51], [52], [53]]. However, previous screens have primarily focused on host dependency factors for human or avian influenza viruses, while largely overlooking SIV and the critical role of pigs as intermediate adaptive hosts that facilitate the emergence of novel pandemic strains. In this context, our study employed a genome-wide CRISPR knockout (GeCKO) screening in porcine NPTr cells to identify essential host factors for SIV replication. The 39 overlapped candidates across the last four rounds of screening, encompassing several well-established and significant host factors, such as SLC35A1, WDR7, and ATP6AP2, confirming the robustness and reproducibility of our screening process. Among these overlapped candidates, subsequent functional validation confirmed the indispensable role of STK11 in supporting IAV infection, both in vitro and in vivo. Although the anticipated protective benefit was not observed in the in vivo experiments, this may be due to the transient nature of siRNA-mediated gene silencing, which could not be maintained throughout the 14-day study period. Notably, similar antiviral effects to those observed in vitro were evident at earlier time points (Fig. 3E-G). Collectively, our genome-wide GeCKO screening in NPTr cells provides an efficient approach to discover vital host genes required for SIV replication and promising targets for the development of host-directed therapeutics for SIV.

STK11 is a highly conserved serine/threonine kinase involved in metabolism, proliferation, and cell morphology [54], and has also been implicated in viral pathogenesis. It promotes viral latency in Kaposi’s sarcoma-associated herpesvirus (KSHV)-infected cells via AMPK activation [55], while its downregulation by HBV-X protein contributes to tumor progression [33]. STK11 also interacts with DENV-2 NS1 to activate AMPK and induce autophagy, facilitating DENV replication [35,56]. AMPK modulation broadly affects replication of viruses such as HCMV, HSV-1, and HCV [57], suggesting STK11 may act through AMPK-dependent pathways. However, our data show that STK11 knockout suppresses IAV replication independently of AMPK signaling (Fig. S5). In addition, previous work has suggested that STK11 can restrict the early stages of replication for several RNA viruses by interfering with fatty acid biosynthesis [58]. Here, we demonstrated that STK11 facilitates IAV attachment by maintaining the spatial organization of sialic acid clusters, revealing a novel mechanism through which STK11 regulates viral infection. Consistent with this, STK11 deficiency also impaired NDV attachment (Fig. 5F, G), suggesting a broader role for STK11 in modulating viruses that utilize sialic acid receptors. Mechanistically, STK11 maintains the organization of sialic acid clusters through promoting the formation of intracellular actin stress fibers. Formation of intracellular actin stress fibers is predominately regulated by RhoA [59,60]. Our data demonstrated that the absence of STK11 reduced the levels of active GTP-bound RhoA as well as the activation of its downstream effectors, MLC and LIMK, thereby compromising RhoA signaling, consistent with previous studies [[61], [62], [63]]. However, further studies are required to identify the specific direct downstream substrate through which STK11 facilitates RhoA signaling.

IAV attachment primarily relies on multivalent interactions between viral HA and sialic acid receptors [64]. Correspondingly, some host factors modulate IAV attachment by altering sialic acid receptor expression levels, such as SLC35A1, APOE, and B3GAT1 [[25], [26], [27]]. In the present study, we identified STK11 as a novel host factor that positively regulates IAV attachment. Mechanistic investigations revealed that the impaired viral attachment observed upon STK11 knockout was attributable to alterations in the spatial organization of sialic acid clusters, rather than to reduced overall receptor abundance at the cell surface. The previous report indicates that both α-2,3 and α-2,6 sialic acid receptors are organized into clusters resembling finger-like microvillus-like protrusions, which provide a multivalent virus-binding platform [41]. Furthermore, interferon-gamma has been shown to inhibit IAV attachment by decreasing the size of sialic acid clusters without affecting receptor abundance, further corroborating the significance of spatial organizations of sialic acid clusters in IAV attachment [65]. We speculate that the disordered, irregular and interdigitated distribution of sialic acid clusters on WT cell surfaces is more conducive to forming multivalent interactions between viral HA and sialic acid receptors than the orderly distribution pattern observed on STK11 knockout cell surface. Additionally, the structure and modifications of sialic acid-containing glycan on cell membranes are also important determinants for efficient IAV attachment [28,66,67]. However, our current study did not investigate whether STK11 deficiency induces alterations in the structure or modifications of sialic acid-containing glycan on the cell surface, which would be worthwhile to investigate in future studies.

Membrane-associated proteins typically exhibit dynamic clustering at the nanoscale [68,69], which is essential for numerous cellular processes, with the actin cytoskeleton playing a critical role in this process [43,70]. The actin-based membrane skeleton and anchored transmembrane proteins partition the membrane into compartments, confining membrane proteins within these regions [71,72]. Furthermore, interactions between membrane components and actin skeleton drive molecular clustering, particularly in the case of glycosyl-phosphatidylinositol (GPI)-anchored proteins [68,73]. Consistently, our study provides evidence for the crucial role of the actin cytoskeleton in sustaining the spatial organization of sialic acid clusters conducive to IAV binding. We speculated that the conversion of sialic acid clusters from a disordered and interdigitated organization pattern into an ordered organized pattern extending outwardly perpendicular to the cell surface after STK11 knockout could be attributed to a decrease in constraints of glycoproteins and glycolipids carrying the receptors, resulted from a reduction in the intracellular actin skeleton. However, the underlying mechanism linking the intracellular actin skeleton and the spatial arrangement of sialic acid clusters remains obscure, which would be valuable to investigate in future studies.

While STK11′s critical role in IAV infection underscores its potential as an antiviral target, its dual function as a tumor suppressor and master metabolic regulator necessitates carefully designed strategies [74]. Inactivation of STK11 is associated with aggressive tumor phenotypes [75], and systemic inhibition may disrupt AMPK-mediated pathways regulating energy balance, fatty acid oxidation, and autophagy, leading to metabolic and immune dysregulation [39,[76], [77], [78]]. Systemic or prolonged inhibition may increase risks of oncogenesis and other adverse effects. Instead, strategies with precise spatial and temporal control—such as tissue-specific inhibitors or delivery systems restricting inhibition to the respiratory tract—are essential to minimize off-target effects [79]. Moreover, transient inhibition during acute infection could further mitigate adverse outcomes compared with chronic inhibition [80,81]. Future efforts should focus on developing such targeted platforms to harness STK11′s antiviral potential safely.

Conclusion

In summary, we identified a novel IAV attachment regulator by an unbiased porcine genome-wide screening. Mechanistic analyses revealed that STK11 maintains the spatial organization of sialic acid clusters by facilitating the formation of intracellular actin stress fibers, thereby supporting efficient IAV attachment. In vivo studies further confirmed the suppressive effect of STK11 knockdown on IAV replication, underscoring its potential as a broad-spectrum host target for antiviral intervention. Given the conserved structure and available crystallographic data of STK11, we are screening small-molecule libraries to identify selective inhibitors. Future work should prioritize evaluating their efficacy and safety in preclinical models and exploring synergistic combinations with existing antivirals. Such strategies may accelerate the development of host-directed therapeutics against influenza and other emerging viral threats.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary Figure 1.

Fig. S1 Validation of polyclonal knockout cells of candidate genes. (A) DNA sequence alignments of genomic segments containing sgRNA targeting sites of polyclonal knockout cells. (B) Cell viability determination of polyclonal knockout cells by CCK-8 assay (n = 8).

Supplementary Figure 2.

Fig. S2 IAV infection does not alter STK11 expression or localization. (A) The effect of IAV infection on STK11 expression. WT PK-15 cells were infected with HuB H1N1 (MOI = 0.01). The cells were collected at the indicated times for western blot analysis of STK11 and viral NP expression. The ratio of STK11 to GAPDH was calculated by ImageJ software. (B) The effect of IAV infection on STK11 subcellular localization. WT PK-15 cells were infected with HuB H1N1 (MOI = 0.1). The cells were subjected to incubation with viral HA and STK11-specific antibody for confocal analysis at 0 and 8 hpi, scale bar = 10 μm.

Supplementary Figure 3.

Fig. S3 STK11 knockdown impairs IAV replication in human 293T cells. (A) Western blot analysis of STK11 expression in WT and polyclonal STK11 knockdown (STK11-KD) 239T cells. (B) Cell viability determination of WT and STK11-KD 293T cells by CCK-8 assay (n = 8). (C) The effect of STK11 knockdown on PR8 H1N1 replication. WT and STK11-KD 293T cells were infected with PR8 H1N1 (MOI = 0.01), and the supernatants were collected at 12, 24, 36 hpi, followed by virus titers determination by TCID₅₀ assay (n = 3). Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (ns, not significant; *, p < 0.05; ****, p < 0.0001).

Supplementary Figure 4.

Fig. S4 The effect of IAV infection on STK11 expression in vivo. Female SPF BALB/c mice were intranasally infected with HuN H1N1 or treated with PBS as a control. The STK11 and viral NP expression in lung homogenates from PBS-treated or HuN H1N1-infected mice at 3 and 5 days post-infection were examined by western blot analysis.

Supplementary Figure 5.

Fig. S5 STK11 supports SIV replication independently of AMPK activation. (A) The effect of STK11 knockout on AMPK phosphorylation. Control and STK11-KO NPTr cells were infected with or without HuB H1N1 (MOI = 0.01). The cells were collected at the indicated times and subjected to western blot analysis. (B) Detection of AMPK phosphorylation in STK11-KO cells. The STK11-KO NPTr cells were treated with different concentrations of 991 for 12 h, and then the cells were subjected to western blot analysis. (C) Cell viability determination of STK11-KO NPTr cells treated with different concentrations of 991 by CCK-8 assay (n = 8). (D) The effect of AMPK reactivation on IAV replication in STK11-KO cells. Control and STK11-KO NPTr cells were treated with or without 50 μM 991 for 12 h, and then infected with HuB H1N1 (MOI = 0.01). The cells were subjected to western blot analysis for viral NP and p-AMPK expression at 24 hpi.

Supplementary Figure 6.

Fig. S6 STK11 is required for efficient IAV attachment. (A, B) Detection of viral NP expression and subcellular distribution. Control and STK11-KO NPTr cells were infected with HuB H1N1 (MOI = 0.1), and the NP expression and distribution were detected by (A) confocal microscopy, scale bar = 50 μm, and (B) western blot assay at 3, 6, 9 hpi. The ratio of NP to GAPDH was calculated using the ImageJ software. (C) Detection of viral M1 expression. Control and STK11-KO NPTr cells were infected with HuB H1N1 (MOI = 10), and the M1 expression was detected by western blot assay from 0.5 to 2.5 hpi. The ratio of M1 to GAPDH was calculated using the ImageJ software. (D) Detection of viral NP expression after acid bypass treatment. Control and STK11-KO NPTr cells were infected with HuB H1N1 (MOI = 0.1) and treated with acid bypass (pH = 5.0) or not (pH = 7.4). The viral NP expression was determined by western blot analysis at 8 hpi. (E) Removal of uninternalized virus via trypsin-EDTA treatment. PK-15 WT cells were infected with HuB H1N1 (MOI = 100) on ice for 1 h. Then the cells were treated with or without 0.25% Trypsin-EDTA, followed by western blot analysis for viral protein expression. (F, G) Colocalization of viral protein with CLTC. (F) WT and STK11-KO PK-15 cells were infected with PR8 H1N1 (MOI = 10) or 115 H9N2 (MOI = 10), and the distribution of NP (PR8) or HA (115) protein and CLTC was visualized by confocal microscopy at 15 min post-infection. Scale bar = 5 μm. (G) The colocalization between viral NP or HA and CLTC in (F) was quantified by calculating Mander’s colocalization coefficients using ImageJ software on 15 fields per sample (n = 15). (H, I) Detection of viral attachment. (H) WT and STK11-KO PK-15 cells were infected with PR8 H1N1 (MOI = 100) or 115 H9N2 (MOI = 100) for 1 h on ice, and then the cells were incubated with NP (PR8) or HA (115) specific antibody, respectively, for confocal analysis, scale bar = 20 μm. (I) The average fluorescence intensity per cell in (H) was calculated by ImageJ software on 10 fields per sample (n = 10). (J) Back-complementation of STK11 restored viral attachment. WT, STK11-KO, and STK11 revertant PK-15 cells were infected with HuB H1N1 (MOI = 100) for 1 h on ice, and the cells were analyzed by western blot for viral protein expression. Data are presented as mean ± Standard Deviation (SD) of independent biological replicates. Statistical analysis was performed using an unpaired, two-tailed Student’s t-test. (****, p < 0.0001).

Supplementary Figure 7.

Fig. S7 The influence of Jasplakinolide treatment on cell viability of STK11-KO cells. STK11-KO PK-15 cells were treated with either DMSO or Jasplakinolide for the indicated durations, followed by cell viability determination by CCK-8 assay (n = 8).