miR-17-5p inhibits classical swine fever virus replication by targeting the PKD2-regulated AMPK/mTOR autophagy pathway

ABSTRACT

Classical swine fever virus (CSFV) remains a major threat to the global swine industry, yet the involvement of host miRNAs in its pathogenic mechanisms is not fully understood. In this study, we demonstrate for the first time that miR-17-5p inhibits CSFV replication through an autophagy-dependent mechanism by targeting polycystin-2 (PKD2), a key calcium channel protein that regulates the AMPK/mTOR signaling pathway. Using the PK-15 cell model, we found that CSFV infection significantly upregulates miR-17-5p expression. Functional assays revealed that miR-17-5p exerts antiviral effects by directly binding to the 3’-UTR of PKD2, as confirmed by bioinformatics prediction and dual-luciferase reporter assays. Silencing of PKD2 recapitulated the antiviral effect of miR-17-5p overexpression, while PKD2 reconstitution restored viral replication by activating AMPK signaling and suppressing mTOR activity, thereby significantly enhancing autophagic flux – as evidenced by increased LC3-II/I ratio and decreased p62 levels. Mechanistically, PKD2 regulates intracellular calcium dynamics, modulating the AMPK/mTOR-autophagy axis to promote CSFV proliferation. This work uncovers a novel host antiviral mechanism in which a miRNA controls virus-induced autophagy via calcium signaling. To our knowledge, this is the first report to establish the pivotal role of miRNA-mediated calcium signaling modulation in flavivirus-host interactions. These findings provide a mechanistic framework and potential therapeutic targets for anti-CSFV interventions focused on PKD2 or autophagy regulation.

Introduction

Classical swine fever (CSF) is a highly contagious viral disease of swine caused by the classical swine fever virus (CSFV), characterized clinically by high fever, leukopenia, and splenic hemorrhagic infarction, and remains a major threat to the global pig industry [1]. CSFV is a member of the Pestivirus genus within the Flaviviridae family, CSFV and possesses a single-stranded positive-sense RNA genome that encodes a large polyprotein. This polyprotein is subsequently cleaved into four structural proteins (C, Erns, E1, and E2) and eight non-structural proteins, including NS3 and NS5B [2]. Although vaccination programs have been widely implemented, the emergence and continued circulation of the sub-genotype 2.1d strain in China has been shown to reduce the protective efficacy of the traditional C strain vaccine [3]. Increasing evidence indicates that CSFV facilitates its replication by hijacking host metabolic pathways such as lipid metabolism [4], amino acid metabolism [5], and glycolysis pathways [6]. Furthermore, the virus exploits the host autophagy machinery to package viral particles, evade immune detection, and suppress the interferon-mediated antiviral response [7]. However, the precise molecular mechanisms governing these virus – host interactions remain incompletely understood.

MicroRNAs (miRNAs) [8] are a class of small, non-coding RNAs typically 19–25 nucleotides in length that regulate gene expression at the post-transcriptional level by binding to complementary sequences on target host or viral mRNAs [9]. Among these, the miR-17 ~ 92 cluster is a highly conserved polycistronic gene cluster in vertebrates that encodes six mature microRNAs: miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1, and miR-92a-1 [10]. This cluster plays critical roles in modulating key cellular processes including the cell cycle, proliferation, and apoptosis [11]. Emerging evidence suggests that that Flaviviridae family exploit host miRNAs, particularly those within the miR-17–92 cluster, to support their replication. For instance, hepatitis C virus (HCV) infection significantly upregulates host miR-17-5p expression [12]. Likewise, Zika virus (ZIKV) leverages host miR-17-5p to enhance its replication, likely through modulation of the endoplasmic reticulum unfolded protein response (UPR), thereby optimizing the intracellular environment [13]. In the case of bovine viral diarrhea virus (BVDV), efficient viral translation and replication require the binding of miR-17 to the S2 site within its 3’-UTR [14]. In contrast, certain miRNAs from this cluster also exert antiviral effects. For instance, both miR-17-5p and miR-20a have been shown to suppress HIV replication by targeting the transcriptional cofactor PCAF, which interacts with the HIV Tat protein [9]. Additionally, miR-17-5p has been implicated in the regulation of Epstein-Barr virus (EBV) latency, where it targets signaling molecules such as SOS1 in the B-cell receptor (BCR) pathway, inhibiting NF-κB activation and plasma cell differentiation to prevent lytic reactivation [15]. These findings highlight the central role of miR-17-5p, a core member of the miR-17 ~ 92 cluster, in modulating the life cycle of various viruses within the Flaviviridae family.Moreover, existing studies have established that miR-17-5p plays a critical regulatory role in diseases linked to non-Flaviviridae pathogens. For example, during Mycobacterium infection, the elevated expression of miR-17-5p can influence host immune regulation and bacterial pathogenicity by downregulating inflammatory mediators [16]. In a model of infectious bursal disease virus (IBDV) infection in chickens, this miRNA targets and modulates the expression of key immune-related genes, thereby indirectly impacting the host’s immune response to the virus [17]. These findings clearly illustrate that microRNAs exhibit both distinct universality and specificity in virus-host interactions.

However, the specific function of miR-17-5p in the context of classical swine fever virus (CSFV) infection has not yet been elucidated.

Autophagy plays a complex and dual regulatory role in viral pathogenesis, acting both as a host defense mechanism that restricts viral infection and as a cellular process that can be subverted by viruses to facilitate their replication. In the case of CSFV, the nonstructural-protein NS5A [18], has been shown to induce autophagy by inhibiting mTOR phosphorylation, thereby promoting the formation of autophagic membranes that serve as scaffold for viral assembly [19]. Similarly, enteroviruses reshape manipulate autophagic trafficking by modulating host tethering proteins to enhance viral replication [20], while in transmissible gastroenteritis virus infection, autophagy has been found to exert antiviral effects by limiting viral proliferation [21]. Importantly, calcium ions (Ca^{2 +}), as a central intracellular second messenger, play a critical regulatory role in autophagy through modulation of the AMPK/mTOR pathway [22]. Calcium homeostasis is maintained by a tightly regulated system within the endoplasmic reticulum (ER), involving polycystin-2 (PKD2)-mediated Ca²⁺ release and its reuptake via the sarco/endoplasmic reticulum Ca²⁺ -ATPase (SERCA) pump [23]. Several viruses have evolved strategies to exploit this calcium-autophagy regulatory axis. For example, porcine reproductive and respiratory syndrome virus (PRRSV) activates CaMKII-dependent autophagy by inducing intracellular Ca²⁺ imbalance [24], and porcine circovirus type 2 (PCV2) relies on CaMKKβ-mediated autophagy, which can be effectively blocked by inhibition of Ca^{2 +} signaling [25]. Rotavirus (RV) utilizes its NSP4 viroporin to deplete ER Ca²⁺ stores, initiating an ER stress-induced autophagy cascade that supports viral replication [26]. This mechanism of leveraging calcium-regulated autophagy appears to be conserved across the Flaviviridae family. For example, Japanese encephalitis virus (JEV) activates inflammatory responses via NMDA receptor-mediated Ca²⁺ influx [27], while ZIKV triggers apoptosis by disrupting ER calcium homeostasis [28]. Together, these findings highlight the calcium-autophagy axis as a key regulatory node in virus – host interactions and suggest it may represent a common target for flaviviral pathogenesis and immune evasion.

Polycystin-2 (PKD2), also known as transient receptor potential cation channel subfamily V member 1 (TRPV1), is a critical calcium-permeable channel that plays a fundamental role in maintaining intracellular Ca²⁺ homeostasis. Previous studies have demonstrated that PKD2 directly interacts with the core autophagy protein Beclin1 to facilitate autophagosome formation [29]. As a member of the transient receptor potential (TRP) channel family, PKD2 exhibits subcellular localization- specific functions: at the plasma membrane, it forms heterodimeric complexes with PKD1 to mediate extracellular signal transduction and calcium influx; at the ER membrane, it acts as a calcium-release channel, regulating sarcoplasmic reticulum calcium stores [23]; and in primary cilia, its activity is negatively regulated through interaction with the membrane protein TACAN [30]. Importantly, accumulating evidence indicates that dysregulation of PKD2 disrupts calcium signaling and impairs autophagy via the AMPK/mTOR signaling axis, thereby contributing to the pathogenesis of diseases such as autosomal dominant polycystic kidney disease and other calcium-related disorders [31]. These findings collectively underscore PKD2 as a central node in calcium-dependent autophagy regulation. However, its role in CSFV infection remains largely undefined. Our previous bioinformatics analysis identified PKD2 as a putative target of miR-17-5p. Notably, PKD2 expression was significantly altered during CSFV infection. On the basis of these observations, we propose a novel hypothesis: a “miRNA-ion channel-autophagy” regulatory axis may be involved in modulating host-virus interactions.The current study is designed to address two key questions: (1) Does CSFV infection upregulates miR-17-5p expression to modulate viral replication? (2) Does miR-17-5p exerts its regulatory effects by targeting PKD2, thereby altering Ca²⁺ dependent AMPK/mTOR signaling and autophagic activity, ultimately influencing CSFV replication? Elucidation of this molecular mechanism will not only advance our understanding of host-directed regulatory networks during CSFV infection but may also identify novel therapeutic targets for the development of effective antiviral strategies.

Materials and methods

Cells and viruses

The porcine kidney cell line PK-15 (ATCC CCL-33) and the CSFV strain (TCID₅₀ = 10 ~ 6.2/0.1 mL) were obtained and propagated according to previously established protocols [32,33]. The PK-15 cells were tested to be mycoplasma free and cultured in a 5% CO₂ humidified incubator at 37°C in Dulbecco’s modified Eagle medium (DMEM, Gibco, USA) supplemented with 10% heat-inactivated newborn calf serum (NCS, Gibco, pH 7.4), 2 mmL-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin (Bioind). The CSFV strain used in this study was a generous gift from Prof. Yongke Sun, Yunnan Agricultural University and was propagated in PK- 15 cells. The multiplicity of infection (MOI) was calculated based on the virus titer (TCID₅₀ = 10 ~ 6.2/0.1 mL) and the cell number per well upon cell seeding. The human embryonic kidney cell line HEK-293T (ATCC CRL-3216) was cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Israel) at 37°C under 5% CO₂. For infection experiments, CSFV was inoculated at a multiplicity of infection (MOI) of 1.

Oligonucleotides and plasmid transfection

The miR-17-5p mimics, inhibitor, and corresponding negative control (miR-NC) were synthesized by RiboBio (Guangzhou, China; sequences listed in Table 1). PK-15 cells at 70–80% confluency were transfected with miRNAs at final concentrations ranging from 0 to 150 nM using Lipofectamine™ 3000 (Thermo Fisher Scientific), following the manufacturer’s protocol. At 12 h post-transfection, cells were either infected with CSFV at an MOI of 1 or subjected to additional treatments, as specified. Two siRNAs targeting porcine PKD2 (designated siPKD2-1 and siPKD2-2) were designed using the BLOCK-iT™ RNAi Designer and synthesized with HPLC-purified (Takara, China; sequences provided in Table 1). Transfection was carried out under the same conditions as miRNA transfection, using a final siRNA concentration of 50 nM. The full-length coding sequences (CDS) of porcine PKD2, MED12L, and FGD4 were cloned into the pcDNA3.1(+) expression vector (Thermo Fisher Scientific) to generate overexpression constructs: pcDNA3.1-PKD2-HA, pcDNA3.1-MED12L-HA, and pcDNA3.1-FGD4-HA. Primer sequences used for cloning are listed in Table 2. All plasmid constructs were sequence-verified and transfected into PK-15 cells (3000 ng/well in six-well plates) using Lipofectamine™ 3000. At 12 hours post-transfection, cells were either infected with CSFV (MOI = 1) or subjected to further experimental procedures.

Table 1.

RT qPCR Primer Information,Simulated substances, inhibitors, and small interfering RNA sequences.

Primer name	Primer sequences (5“→3”)	Reference sequence
miR-17-5p-loop	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACCTACCT	MIMAT0007755
miR-17-5p-F	GCGCAAAGTGCTTACAGTGC
miR-17-5p-R	AGTGCAGGGTCCGAGGTATT
mimics	CAAAGUGCUUACAGUGCAGGUAG
mimicsNC	GUAACCUCCCGGAAAUCCUGC
mimics	CAAAGUGCUUACAGUGCAGGUAG
mimicsNC	GUAACCUCCCGGAAAUCCUGC
u6-F	AGTTCCGGGTACAATACGCC	NR-004394.1
u6-R	TTACTGTCACGCTCGGTCAC
CSFV-F	CGTGGCAGGTTCCTTTAAAG	KJ661548
CSFV-R	TTCCCTTGACAACAGGACTC
GAPDH-F	TGACAACTCCCTCAAGATCG	XM-021091114.1
GAPDH-R	AAGCAGGGATGATGTTCTGG
PKD2-F	ACGGGCTCTACTGGAAGAC	NM_001245979.1
PKD2-R	CTCTTTAATCTCATCTCTCAAGTCCTG
siPKD2-1	CAAGAUUGAUGCAGUGAUAGU
	ACUAUCACUGCAUCAAUCUUG
siPKD2-2	CUAUGUGGUGGAAGAGAUACU
	AGUAUCUCUUCCACCACAUAG
siNC	TTCTCCGAACGTGTCACGT	Random sequence
MED12L-F	GAAGGATTGTGTGGTGTGGT	XM_021069618.1
MED12L-R	ACAGGCACTACTGAAGAGGT
FGD4-F	GCTTTGCTGCAGACAGTGA	XM_021092072.1
FGD4-R	CCAAAGTGTCCAGATGCTTGC
PRRG1-F	CGTGTAGCATTGAGAACT	NC_010461.5
PRRG1-R	CAAAGGTGTGGCCCTA
UEVLD-F	AACTGACAGACTTCC	NC_010444.4
UEVLD-R	CTGTGAGCTGTGGTGTA
VANGL1-F	ATGACAACTG GGGCGAGA	XM_013997257.2
VANGL1-R	AGATGCGGACCCCGTAGAAA

Table 2.

Primer information.

Primer name	Primer sequences (5“→3”)	Amplicon length	Reference sequence
PKD2-F1	GCCACCATGGTGAACTCGAGCCGCGTG	2913bp	NM_001245979.1
PKD2-R1	AACATCGTATGGGTACACGTGGATATCG
PKD2-F2	GCCACCATGGTGAACTCGAGCCGCGTG
PKD2-R2	CTAAGCGTAATCTGGAACATCGTATGGGTACAC
MED12L-F1	GCCACCATGGCTTCCAGCACAGGCGATG	5979bp	XM_021069618.1
MED12L-R1	GATTCTCAATAAAATCCATCATGAAG
MED12L-F2	CTTCATGATGGATTTTATTGAGAATC
MED12L-R2	AACATCGTATGGGTAGAAGTGTGAAGG
MED12L-F3	CTTCATGATGGATTTTATTGAGAATC
MED12L-R3	CTAAGCGTAATCTGGAACATCGTATGGGTAGAA
FGD4-F1	GCCACCATGGAGGAAGTAAAACCTGCCT	2301bp	XM_021092072.1
FGD4-R1	ACATCGTATGGGTAGCATTCTGATTTTC
FGD4-F2	GCCACCATGGAGGAAGTAAAACCTGCCT
FGD4-R2	CTAAGCGTAATCTGGAACATCGTATGGGTAGCA

miRNA target prediction

Putative target genes of miR-17-5p were initially identified using two established bioinformatics platforms: TargetScan (http://www.targetscan.org/vert_80/) and miRDB (https://mirdb.org). A total of 936 overlapping candidate genes were retrieved for further experimental validation. Candidate selection was based on the following criteria: (1) Binding free energy of miRNA-3’UTR interaction ≤ −20 kcal/mol; (2) Phylogenetic conservation of the miRNA seed region across vertebrate species; and (3) Perfect Watson-Crick base pairing within the seed sequence of miR-17-5p.

RNA extraction and quantitative real-time PCR (qRT-PCR)analyses

Total RNA was extracted from PK-15 cells – either untreated or subjected to CSFV infection and/or transfection – using the RNAiso Plus RNA isolation kit (Takara, Dalian, China; Cat. #9109), following the manufacturer’s instructions, as described previously [31–34]. total RNA was isolated from PK-15 cells without or with treatments (infection or/and transfection)using the RNAiso Plus RNA isolation kit(Takara Dalian, China, Cat. #9109). Reverse transcription was performed using the PrimeScript RT Reagent Kit (Perfect Real Time, Takara Dalian, China, Cat. No. RR037A). Quantitative PCR (qPCR) was conducted using SsoFast™ Eva-Green® Supermix (Bio-Rad, Cat. No. 172–5201AP) on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, USA). Primer sequences are listed in Table 1.

Relative gene expression levels of PKD2, MED12L, FGD4, PRRG1, UEVLD, VANGL1, and CSFV E2 were determined using the 2^−ΔΔCt method, as described in ABI PRISM 7700 Sequence Detection System user Bulletin # 2 (1999). GAPDH served as the internal control for normalization. For absolute quantification of CSFV, the viral E2 gene was cloned into the pMD18-T vector (Takara, Cat. No. 6011, Dalian, China), transformed into E.coli for bulk culture, and confirmed by sequencing. The recombinant plasmid was serially diluted 10-fold to generate a standard curve for quantifying the absolute copy number of CSFV E2 gene RNA in infected PK-15 cells. For miRNA amplification, cDNA synthesis was synthesized using the miRCute miRNA First-Strand Synthesis Kit (Vazyme, Cat.#MR101-01), and qRT-PCR was carried out using the miRNA Universal SYBR qPCR Master Mix (Vazyme, Cat.#MQ101-01). U6 snRNA served as the endogenous control. All miRNA primers are listed in Table 1.

Western blot analysis

As previously described [32,34–36], total protein was extracted from treated or control PK-15 cells using RIPA lysis buffer (Beyotime Biotechnology, China; Cat. #P0013B) supplemented with protease and phosphatase inhibitors. Protein concentrations were determined using a BCA protein assay kit (Beyotime; Cat. #P0010), following the manufacturer’s instructions. Equal amounts of total protein (10 μg per lane) were resolved SDS-PAGE on either 6% or 10% Bis-Tris NuPAGE gels (Thermo Fisher Scientific), depending on the molecular weight of the target protein. Following electrophoresis, proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad; Cat. #1620177) at a constant current of 180 mA for 1 h. Membranes were blocked with 5% skim milk in TBST for 1 h at room temperature and then incubated overnight at 4°C with the appropriate primary antibodies (Table 3) under gentle agitation. After three 10-minute washes with TBST, membranes were incubated for 1 h at room temperature with corresponding horseradish peroxidase (HRP)-conjugated secondary antibodies (Table 3). Protein bands were visualized using ECL Plus chemiluminescent substrate (Bio-Rad; Cat. #1705060) and captured using a TOUCH IMAGER chemiluminescence imaging system (Fusion FX). Band intensities were quantified using ImageJ software (version 1.5) and normalized to GAPDH, which served as the internal loading control.

Table 3.

Antibody information.

Antibodies	Vendors	cat.No.
Anti GAPDH mouse mAb	Proteintech	60,004–1-Ig
Anti CSFV E2 mouse mAb	Biorbyt	Or318086
Anti PKD2 rabbit mAb	Nature	A17778
Anti mTOR mouse mAb	proteintech	66,888–1-lg
Anti AMPKrabbit PolyAb	proteintech	10,929–2-AP
Anti pmTOR mouse mAb	proteintech	67,778–1-lg
Anti pAMPK rabbit mAb	Cell signaling	2535T
Anti LC3B rabbit mAb	Nature	A27936
Anti P62rabbit PolyAb	proteintech	18,420–1-AP
Anti CSFV E2 (IF)mouse mAb	Qianxun Biotechnology Co., Ltd.	Ab-025
HRP-GoatAnti-MouseIgG(H+L)	Proteintech	SA00001-1
HRP-GoatAnti-RabbitIgG(H+L)		SA00001-2

Luciferase reporter assay

Potential miR-17-5p binding sites within the 3”-UTR of the porcine PKD2 gene were predicted using TargetScan, miRDB, and RNAhybrid, identifying three candidate binding sites:Site 1 (173–179 bp); Site 2 (563–570 bp) and Site 3 (2824–2831 bp). To functionally validate these interactions, wild-type (WT) and mutant (MUT) fragments of the PKD2 3”-UTR (with Site 2 mutated from GCACTTTA to ATGTCCCG) were synthesized and cloned downstream of the XhoI/XbaI restriction sites into the pGL3-control luciferase reporter vector (Promega, USA).

HEK-293T cells, seeded in 24-well plates at 70–80% confluency, were co-transfected with 100 ng of WT or MUT pGL3 reporter plasmid and 50 nM of miR-17-5p mimics or negative control (miR-NC) and pGL3 reporter plasmids using Lipofectamine™ 3000 (Thermo Fisher Scientific). After 36 h, luciferase activity was measured using the Dual-Luciferase Reporter Assay Kit (Vazyme, Cat. #DL101-01), following the manufacturer’s protocol. Firefly luciferase activity was normalized to Renilla luciferase activity to control for transfection efficiency, and results were expressed as relative luciferase activity, with the miR-NC group set as 1. A significant reduction in relative luciferase activity (p < 0.05) in the miR-17-5p group compared to the control was considered indicative of a functional interaction between miRNA and 3’-UTR.

Calcium flux assay

To assess intracellular calcium dynamics, PK-15 cells were transfected with siPKD2-2 (0–100 nM) for 24 h, followed by loading with Fura-2 AM (5 μM; GIVEI, #G&VS10152.1) for 60 min at 37°C in the dark. After incubation, cells were washed with HBSS containing 3 mM EGTA (Yeasen, Cat. #60339ES10) to eliminate extracellular Ca^{2 +}, and baseline Ca^{2 +} levels were recorded for 1 min at 10 s intervals. Cells were then stimulated with one of the following agents to evoke intracellular calcium transients: ATP (100 μM; Aladdin, Cat. #A759861), Thapsigargin (1 μM; MCE, Cat. #HY-13433), or Ionomycin (10 μM; MCE, Cat. #HY-13434). Fluorescence intensity was monitored for an additional 10 min at 10 s intervals using a Fura-2-compatible microplate reader (excitation wavelengths: 340/380 nm; emission: 510 nm). The F340/F380 fluorescence ratio was calculated to indicate intracellular Ca^{2 +} concentrations. Peak amplitude data were extracted and analyzed using GraphPad Prism software.

Autophagy assays

Autophagic activity was assessed by Western blot under various pharmacological and genetic conditions. Briefly, PK-15 cells were subjected to the following treatments: BAPTA-AM (10 μM; Sigma-Aldrich, Cat. #A1076) for 6 h to chelate intracellular Ca^{2 +}; siPKD2-2 (100 nM) transfection for 12 h; Sapanisertib (100 nM; MCE, Cat. #HY-13373) for 6 h to inhibit mTOR signaling; Compound C (10 μM; MCE, Cat. #HY-13418A) for 6 h as an AMPK inhibitor; miR-17-5p mimics (100 nM) transfection for 12 h; Co-transfection with miR-17-5p mimics and pcDNA3.1-PKD2 plasmid for 12 h.

In some selected groups, CSFV infection was performed at an MOI of 1 following treatment, and cells were harvested at specific time points (e.g. 24 and 48 hpi). Autophagy markers were analyzed by Western blotting, including the LC3-II/LC3-I ratio as an indicator of autophagosome formation, p62/GAPDH to assess autophagic flux, phosphorylated AMPK (Thr172) and mTOR (Ser2448) to monitor upstream signaling, normalized to their respective total protein levels, CSFV E2 protein expression to evaluate viral replication.

Virological assays

Viral RNA levels and protein expression were quantified using RT-qPCR and Western blot, respectively, as previously described. For infectivity titration, PK-15 cells were seeded in 96-well plates and inoculated with 10-fold serial serial dilutions of CSFV (100 μL/well, 8 replicates per dilution). Control wells containing uninfected cells were included. After 72 h of incubation at 37°C with 5% CO₂, cytopathic effects (CPE) were evaluated under an inverted microscope. The 50% tissue culture infectious dose (TCID₅₀) was calculated using the Reed-Muench method and expressed as Log₁₀ TCID₅₀/0.1 mL.

The indirect immunofluorescence assay was performed with modifications according to previously described methods [32,33]. Briefly, CSFV-infected cells were washed three times with PBS and fixed with pre-chilled methanol at −20°C for 15 min. After three additional PBS washes (5 min each), cells were blocked with 5% skim milk in PBS for 2 h at room temperature. Primary antibodies (Table 3) were applied overnight at 4°C, followed by five PBS washes (5 min each). Appropriate secondary antibodies (Table 3) were added for 1 h at 37°C in the dark. After another five PBS washes, nuclei were counterstained with DAPI for 15 min at room temperature. Fluorescent images were captured using an Olympus inverted fluorescence microscope equipped with appropriate filter sets.

Cell viability analysis

Carry out experiments based on the previously formulated plan [37]. The PK-15 cells (10⁴ cells/well in 100 μL) were seeded into 96-well plates and incubated for overnight. Transfection with the indicated concentrations of miR-17-5p mimic (0, 25, 50, 100, 150 and 200 nM) was performed as described above, with each treatment in triplicate. Seventy-two hours post transfection, the cell viability analysis using CCK-8 kit (Sigma-Aldrich, Cat. No. 96,992) was performed according to the manufacturer’s guide. In brief, 10 μl of CCK-8 solution was added to each well of the plate. Incubate the plate for 2 h in the incubator and measure the absorbance at 450 nm using a microplate reader. The wells without transfection were used as control for normalization.

Statistical analysis

All statistical analyses and graphical representations were performed using GraphPad Prism software (version 10.1.2). Quantitative data are presented as mean ± standard deviation (SD) from at least three independent experiments. The two-tailed and unpaired Student’s t-test was employed to evaluate the statistical significance of differences in gene expression or protein levels between treatment and control groups (e.g. CSFV infection and/or transfection conditions). Densitometric analysis of Western blot bands was conducted using ImageJ (version 1.5), with protein expression levels normalized to GAPDH. Statistical significance was defined as follows: p < 0.05 (*), statistically significant; p < 0.01 (**), highly significant; p < 0.001 (***) extremely significant.

Results

CSFV infection induces miR-17-5p expression

To assess the dynamic changes in miR-17-5p expression following CSFV infection, RT-qPCR was performed on PK-15 cells collected at various time points post-infection (0–48 hpi). As shown in Figure 1A, CSFV infection significantly upregulated miR-17-5p expression in a time-dependent manner compared to the non-infected control group. The expression level progressively increased with prolonged infection duration. These findings indicate that CSFV infection specifically induces the upregulation of miR-17-5p in host cells.

Figure 1.

CSFV infection upregulates miR-17-5p expression.

(A, B) PK-15 cells were infected with CSFV (MOI = 1), and intracellular levels of miR-17-5p (A) and CSFV RNA (B) were quantified by RT-qPCR at the indicated time points post-infection. Data are presented as mean ± standard deviation (SD) from three independent experiments. Statistical significance was determined using two-tailed Student’s t-tests: ns, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001. Error bars represent SD.

miR-17-5p inhibits CSFV replication

Given the significant upregulation of miR-17-5p in PK-15 cells following CSFV infection, we hypothesized that miR-17-5p may play a functional role during CSFV infection. To investigate this, PK-15 cells were transfected with increasing concentrations (0–150 nM) of miR-17-5p mimics prior to infection with CSFV (MOI = 1). Cells treated with 0 nM mimics received an equivalent volume of DMSO as a vehicle control. After infection, CSFV replication was assessed via IFA, RT-qPCR, and Western blot. As shown in Figure 2A–C, miR-17-5p mimics treatment resulted in a dose-dependent reduction in CSFV E2 fluorescence intensity (IFA), E2 gene RNA level (RT-qPCR), and E2 protein expression (Western blot). Based on the these findings, 100 nM miR-17-5p mimics was selected for use in subsequent experiments. To assess whether miR-17-5p mimetics have an impact on the viability of PK-15 cells, we transfected PK-15 cells with mimetics at various concentrations. The subsequent CCK-8 assay results indicated that miR-17-5p mimetics did not exert a significant effect on cell viability across all tested concentration ranges. This confirms that the observed effects on viral replication are specific and not attributable to cytotoxicity (Figure 2D).

Figure 2.

miR-17-5p inhibits CSFV replication PK-15 cells.

To further evaluate the effect of miR-17-5p on CSFV replication, transfected PK-15 cells were infected with CSFV (MOI = 1), and samples were collected at 0, 24, and 48 hpi, corresponding to the period of active viral replication (Figure 1B) for MCA. As shown in Figure 2E, E2 gene RNA levels were comparable between the miR-NC and Mock groups at both 24 and 48 hpi. In contrast, cells transfected with the miR-17-5p mimics exhibited significantly reduced E2 gene RNA levels compared to the miR-NC group (24 hpi: p < 0.001; 48 hpi: p < 0.01). Supernatant CSFV RNA levels were comparable between the miR-NC and Mock groups at both 24 and 48 hpi. In contrast, cells transfected with the miR-17-5p mimics exhibited significantly reduced Supernatant CSFV RNA levels compared to the miR-NC group (24 hpi: p < 0.01; 48 hpi: p < 0.001, Figure 2F) Consistent with these findings, Western blot analysis revealed that E2 protein expression in the miR-NC group was not significantly different from that in the Mock group at either time point, whereas the miR-17-5p mimics group displayed a marked decrease in E2 protein levels compared to the miR-NC group (Figure 2H).

To evaluate the overall impact of miR-17-5p on CSFV replication, viral titers were determined using the TCID₅₀ assay. PK-15 cells were transfected with 100 nM miR-17-5p mimics and subsequently infected with serial 10-fold dilutions of CSFV. After 3 days of incubation, TCID₅₀ analysis revealed that viral titers in the miR-NC group were comparable to those the Mock group, whereas the miR-17-5p mimics group exhibited a significant reduction in viral titer (p < 0.0001, Figure 2G). Taken together, these results demonstrated that miR-17-5p significantly suppresses CSFV replication in host cells.

(A-C) PK-15 cells were transfected with miR-17-5p mimics (0–150 nM; 0 nM group treated with an equal volume of DMSO as control) for 12 h, followed by CSFV infection (MOI = 1). Intracellular CSFV levels were assessed using indirect immunofluorescence staining (IFA) (green: CSFV E2 protein; blue: nuclei; merged: overlay) (A), RT-qPCR (B), Western blot, and band intensities were quantified using Image J(C). (D) Cell viability of miR-17-5p-transfected PK-15 cells at different concentrations was measured using the CCK-8 assay. (E and F) PK-15 cells were transfected with 100 nM miR-17-5p mimics for 12 h and then infected with CSFV (MOI = 1). Samples were collected at 0-, 24-, and 48-hours post-infection (hpi). Intracellular CSFV E2 gene RNA (E) and extracellular viral RNA in the supernatant (F) were quantified by RT-qPCR. (G) For viral titer analysis, PK-15 cells were transfected with 100 nM miR-17-5p mimics or negative control miR-NC for 12 h, followed by CSFV inoculation using ten-fold serial dilutions. After 3 days, viral titers were determined using the TCID₅₀ assay and expressed as Log₁₀ TCID₅₀/0.1 mL. (H) PK-15 cells were transfected with 100 nM miR-17-5p mimics for 12 h and then infected with CSFV (MOI = 1).CSFV E2 protein expression in cell lysates was examined at the indicated time points (0, 24, and 48 hpi) by Western blot,the relative intensity of each band was quantified through Image J. Data are presented as mean±standard deviation from three independent experiments. Asterisks indicate statistical significance compared to control values (ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; two-tailed Student’s t-test and One-way ANOVA).

Identification of PKD2 as a functional host target of miR-17-5p during CSFV infection

To elucidate the molecular mechanism by which miR-17-5p inhibits CSFV replication, we first employed bioinformatics analyses using TargetScan and miRDB to predict potential host target genes. TargetScan identified 1385 candidate targets, while miRDB yielded 1333 targets, with 936 genes overlapping between the two databases (Figure 3A). Based on stringent criteria – including binding site specificity (8mer > 7mer-m8 > 7mer-A1 > 6mer); binding score (Context++ score ≤ −0.25, Percentile ≥97, Weighted context++ score ≤ −0.25); and conservativeness (PCT ≥0.93), six candidate target genes were selected for further validation: PKD2, MED12L, FGD4, PRRG1, UEVLD, and VANGL1 (Table 4).

Figure 3.

PKD2 is a host target of miR-17-5p involved in the regulation of CSFV replication. (A) venn diagram showing predicted miR-17-5p target genes identified by by TargetScan (1,385 genes) and miRDB (1,333 genes), with 936 overlapping targets. (B) PK-15 cells were transfected with 100 nM miR-17-5p mimics or negative control (miR-NC) for 12h. mRNA levels of candidate target genes were measured by RT-qPCR. (C, D) PK-15 cells were transfected for 12 h with overexpression plasmids for PKD2 (OE-PKD2), MED12L (OE-MED12L), FGD4 (OE-FGD4), or control vector, followed by CSFV infection at MOI=1 for 24 h. CSFV RNA and protein levels were analyzed by RT-qPCR (C) and Western blot, (D),respectively, and band intensities were quantified using image J. (E) for viral titer analysis, PK-15 cells were transfected with PKD2 (OE-PKD2), MED12L (OE-MED12L), or FGD4 (OE-FGD4), followed by CSFV inoculation using ten-fold serial dilutions. After 3 days, viral titers were determined using the TCID₅₀ assay and expressed as Log₁₀ TCID₅₀/0.1mL. (F, G) PK-15 cells were transfected with increasing concentrations of miR-17-5p mimics (0–150 nM) for 12 h. Expression of PKD2 at the mRNA and protein levels was evaluated by RT-qPCR (F) and Western blot (G), respectively,the relative intensity of each band was quantified through image J. Data are presented as mean ± standard deviation (SD) from three independent experiments. Statistical significance was determined using Student’s t-test and one-way ANOVA. ns, not significant; *p < 0.05; **p < 0.01; *** p < 0.001; **** p < 0.0001.

Table 4.

Binding analysis of miR-17-5p potential target genes.

	Site type	Context++ score	Context++ Score percentile	Weighted context++ score	Conserved Branch length	PCT	Predicted Relative KD
PKD2	8mer	−0.33	98	−0.33	5.140	0.93	−5.102
UEVLD	8mer	−0.38	99	−0.32	2.111	<0.1	−5.752
MED12L	8mer	−0.25	97	−0.25	5.762	0.95	−5.851
VANGL1	8mer	−0.31	98	−0.31	8.543	0.96	−4.886
PRRG1	8mer	−0.46	99	−0.46	7.477	0.96	−5.196
FGD4	8mer	−0.32	98	−0.32	5.830	0.95	−5.578

To experimentally validate these predictions, PK-15 cells were transfected with 100 nM miR-17-5p mimics for 12 hours, and mRNA levels of the candidate genes were assessed by RT-qPCR. The results showed that PKD2 (p < 0.01), MED12L (p < 0.01), and FGD4 (p < 0.05) were significantly downregulated upon miR-17-5p overexpression (Figure 3B). To assess the functional relevance of these targets in CSFV replication, we overexpressed PKD2, MED12L, and FGD4 in PK-15 cells, followed by CSFV infection. RT-qPCR analysis revealed that only PKD2 overexpression significantly enhanced CSFV E2 gene RNA accumulation (p < 0.001), whereas MED12L and FGD4 had no significant effect (Figure 3C). Western blot analysis revealed that only PKD2 overexpression significantly enhanced E2 protein levels accumulation (p < 0.01), whereas MED12L and FGD4 had no significant effect (Figure 3D). To evaluate the overall impact of host target genes on CSFV replication, viral titers were determined using the TCID₅₀ assay. PK-15 cells were transfected with PKD2, MED12L, or FGD4 and subsequently infected with serial 10-fold dilutions of CSFV. After 3 days of incubation, TCID₅₀ analysis revealed that viral titers in the PKD2 group exhibited a significant reduction in viral titer (p < 0.0001, Figure 3E). These findings suggests that PKD2 is a critical host factor mediating the antiviral activity of miR-17-5p.

To further confirm this regulatory relationship, we examined PKD2 expression in PK-15 cells transfected with increasing concentrations of miR-17-5p mimics (50, 100, and 150 nM). Both RT-qPCR (Figure 3F) and Western blot (Figure 3G) analyses demonstrated a dose-dependent reduction in PKD2 expression at the mRNA and protein levels, confirming that miR-17-5p negatively regulates PKD2. Collectively, these data identify PKD2 as a key functional target through which miR-17-5p exerts its inhibitory effect on CSFV infection.

miR-17-5p targets the 3’-UTR of PKD2 mRNA

To further validate the regulatory relationship between miR-17-5p and PKD2, we employed multiple bioinformatics tools (TargetScan, RNAhybrid, miRDB, and PicTar) to predict potential miR-17-5p binding sites within the 3’-untranslated region (3’-UTR) of PKD2 mRNA. Three candidate binding sites were identified: site 1 (173–179 bp; Figure 4A), site 2 (563–570 bp; Figure 4B), and site 3 (2854–2831 bp; Figure 4C). To experimentally confirm the direct interaction between miR-17-5p and these predicted sites, luciferase reporter constructs (pGL3-PKD2-3”-UTR-WT1/MUT1, WT2/MUT2, WT3/MUT3) containing wild-type (WT) or mutated (MUT) versions of the respective sequences were generated. These plasmids were co-transfected into HEK-293T cells along with either miR-17-5p mimics or negative control (miR-NC). Luciferase assay results (Figure 4D) demonstrated that miR-17-5p mimics significantly reduced the luciferase activity containing WT site 2 (WT2) (p < 0.01), whereas no significant changes were observed for WT1 or WT3 constructs (p > 0.05). Furthermore, mutation of the seed sequence in site 2 (MUT2: GCACTTTA → ATGTCCCG, 563–570 bp) completely abolished this inhibitory effect (Figure 4E) (p < 0.0001). These findings indicate that miR-17-5p directly targets the 563–570 bp site (GCACTTTA) within the 3”-UTR of PKD2 mRNA to mediate post-transcriptional silencing.

Figure 4.

miR-17-5p directly targets the 3’-UTR (563-570bp) of PKD2 mRNA. (A-C) Bioinformatic analysis using TargetScan, RNAhybrid, miRDB, and PicTar identified three putative binding sites for miR-17-5p within the 3’-untranslated region (3’-UTR) of PKD2 3UTR mRNA. Sequence alignment revealed three complementary regions, with the seed sequences of miR-17-5p indicated by underlining, and corresponding mutant sequences presented for comparison. (D, E) HEK-293T cells were co-transfected with either miR-17-5p mimics or negative control (mimics NC), along with luciferase reporter constructs containing either wild-type (PKD2-3’-UTR-WT) or mutant (PKD2-3’-UTR-MUT) sequences. Dual-luciferase reporter assays were performed 36 h post-transfection, and luciferase activity was normalized to renilla luciferase (Firefly/renilla ratio). Data are expressed as mean ± standard deviation, with error bars representing three independent experimental replicates. Statistical significance was determined using Student’s t-test: ns, not significant; **p < 0.01; ****p < 0.0001. Representative blots are shown from three independent experiments.

Small interfering RNA-mediated knockdown of PKD2 suppresses CSFV replication

To elucidate the functional role of PKD2 in CSFV infection, two specific siRNAs, designated siPKD2-1 and siPKD2-2, were designed targeting porcine PKD2 using the BLOCK-iT™ RNAi Designer tool. These siRNAs were transfected into PK-15 cells, and knockdown efficiency was assessed at 24 h and 48 h post-transfection by RT-qPCR. Compared to the negative control (siNC), siPKD2-1 reduced PKD2 mRNA expression by 57.89% (24 h: p < 0.01; 48 h: p < 0.001), while siPKD2-2 achieved a more effective knockdown of 73.02% (24 h: p < 0.001; 48 h: p < 0.0001) (Figure 5A), confirming effective silencing of PKD2, with siPKD2-2 showing superior efficacy. Subsequently, the transfected cells were infected with CSFV at MOI = 1. RT-qPCR analysis at 24 h and 48 h post-infection revealed that CSFV E2 gene RNA levels were significantly reduced in both knockdown groups. Specifically, siPKD2-1 resulted in a 56.04% decrease (p < 0.01 at 24 h; p < 0.001 at 48 h), while siPKD2-2 led to 69.86% reduction (p < 0.001 at 24 h; p < 0.0001 at 48 h) (Figure 5A). To evaluate the overall impact of two specific siPKD2 on CSFV replication, viral titers were determined using the TCID₅₀ assay. PK-15 cells were transfected with siNC, siPKD2-1, or siPKD2-2 and subsequently infected with serial 10-fold dilutions of CSFV. After 3 days of incubation, TCID₅₀ analysis revealed that viral titers in the siPKD2-1 and siPKD2-2 group exhibited a significant reduction in viral titer (p < 0.0001, Figure 5B). Western blot analysis further confirmed that protein levels of both PKD2 and CSFV E2 were markedly diminished in the siPKD2-treated group (Figure 5C), consistent with the knockdown efficiencies observed at the mRNA level. Collectively, these results demonstrate that silencing of PKD2 effectively inhibits CSFV replication, highlighting PKD2 as a proviral host factor.

Figure 5.

Down-regulation of PKD2 inhibits CSFV replication. (A) PK-15 cells were transfected with siNC, siPKD2-1, or siPKD2-2, followed by infection with CSFV at an MOI of 1 after 12 h. Intracellular levels of PKD2 mRNA and CSFV RNA were quantified by RT-qPCR at 0, 24, and 48 hr post-infection. (B) for viral titer analysis, PK-15 cells were transfected with siNC, siPKD2-1, or siPKD2-2, followed by CSFV inoculation using ten-fold serial dilutions. After 3 days, viral titers were determined using the TCID₅₀ assay and expressed as Log₁₀ TCID₅₀/0.1mL.(C) Western blot analysis was performed to assess the expression levels of PKD2 and CSFV E2 at the corresponding time points, and band intensities were quantified using image J. (D) PK-15 cells were transfected with a miR-17-5p mimics or co-transfected with the miR-17-5p mimics and a PKD2 overexpression plasmid (OE-PKD2), followed by CSFV infection (MOI = 1) after 12 h. Intracellular CSFV E2 protein levels were analyzed by Western blot, and band intensities were quantified using image J. (E) PK-15 cells were transfected with a miR-17-5p inhibitor or co-transfected with the inhibitor and siPKD2-2, followed by CSFV infection (MOI=1) after 12 h. CSFV E2 protein expression was examined by Western blot (upper panel), and the relative grayscale intensities were quantified by image J (lower panel). Data are presented as mean ± standard deviation (SD) from three independent experiments. Error bars represent SD. Asterisks denote statistical significance compared with the siNC group (A, B) or Mock group (C, D): ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 (Student’s t-test and one-way ANOVA). Representative blot images from three independent experiments are shown.

To determine whether PKD2 functions as a critical downstream effector of miR-17-5p in regulating CSFV replication, backfill experiments were conducted. PK-15 cells were first transfected with miR-17-5p mimics or co-transfected with miR-17-5p mimics and PKD2 overexpression plasmid (OE-PKD2), followed by infection with CSFV at MOI = 1 after 12 hr. An untransfected group infected only with CSFV (Mock) served as the control. Western blot analysis revealed that CSFV E2 protein levels were significantly decreased in the miR-17-5p mimics group compared with the Mock group. Notably, co-transfection of miR-17-5p mimics and OE-PKD2 significantly reversed this inhibition, restoring E2 protein expression to levels higher than those observed with miR-17-5p mimics alone (Figure 5D). In a complementary experiment, PK-15 cells were transfected with a miR-17-5p inhibitor or co-transfected with both the inhibitor and siPKD2-2, followed by CSFV infection (MOI = 1) after 12 h. Compared to the Mock group, miR-17-5p inhibition markedly increased E2 protein expression. However, this effect was significantly attenuated by simultaneous knockdown of PKD2 using siPKD2-2 (Figure 5E).

Collectively, these results demonstrate that knockdown of PKD2 significantly suppresses CSFV replication, with siPKD2-2 exhibiting greater silencing efficiency and antiviral effect. Therefore, siPKD2-2 was chosen for subsequent experiments for functional studies. Moreover, overexpression or knockdown of PKD2 effectively reversed the antiviral or proviral effects of miR-17-5p, respectively, indicating that PKD2 is a functional mediator of miR-17-5p‘s regulatory role in CSFV infection.

PKD2 modulates autophagy through regulation of intracellular Ca^{2 +} signaling

As a member of the transient receptor potential cation channel family, PKD2 is implicated in the regulation of intracellular Ca^{2 +} homeostasis. Given the critical role of Ca^{2 +} signaling in autophagy regulation, we investigated whether PKD2 influences autophagic activity via modulation of Ca^{2 +} dynamics. Intracellular Ca^{2 +} levels were monitored in PK-15 cells using the calcium-sensitive fluorescent probe Fura-2 AM following treatment with gradient concentrations of siPKD2 (0–100 nM) under under conditions of extracellular Ca^{2 +} chelation (3 mM EGTA). Cells were subsequently stimulated with three Ca^{2 +} -mobilizing agents: ATP (10 μM), to induce ER Ca^{2 +} release; thapsigargin (Tg, 1 μM) to inhibit SERCA pumps; and ionomycin (Iono, 1 μM), to mimics store-operated Ca^{2 +} influx. As shown in Figure 6A, siPKD2-treated cells displayed a significant, dose-dependent reduction in intracellular Ca^{2 +} release, as evidenced by a lower F/F₀ fluorescence ratio compared to the siNC control group (*p < 0.05; ** p < 0.01).

Figure 6.

PKD2 regulates cellular autophagy by modulating intracellular affecting Ca²⁺ dynamics. (A) PK-15 cells were transfected with increasing concentrations of siPKD2 (0, 25, 50, 100 nM) for 12 h and then loaded with the calcium-sensitive dye Fura2-AM probe (5 μM). Following chelation of extracellular Ca^{2 +} with 3 mM EGTA, cells were stimulated with ATP (10 μM), thapsigargin (tg, 1 μM), or ionomycin (iono, 10 μM). Real-time intracellular Ca^{2 +} flux was measured and expressed as F/F ₀ fluorescence intensity ratio. (B) PK-15 cells were treated with the intracellular Ca^{2 +} chelator BAPTA-AM (6 h) or transfected with siPKD2-2 (12 h). Western blot analysis was performed to assess LC3-II, LC3-I, and p62 protein levels. The LC3-II/LC3-I ratio and p62 level were quantified using image J. (C) PK-15 cells were transfected with miR-NC, miR-17-5p mimics, or co-transfected with miR-17-5p mimics and a PKD2 overexpression plasmid (OE-PKD2), respectively. Protein levels of LC3-II, LC3-I, and p62 were assessed by Western blot. Data are presented as mean ± standard deviation (SD) from three independent experiments. Asterisks denote statistically significant differences compared to the respective control groups: ns, not significant; *p<0.05; **p<0.01; ***p<0.001 (Student’s t-test and one-way ANOVA). Representative immunoblots from three independent biological replicates are shown.

To further investigate the functional relevance of the PKD2-Ca^{2 +} axis in autophagy, PK-15 cells were treated with either the intracellular Ca^{2 +} chelator BAPTA-AM or siPKD2. Western blot analysis of canonical autophagy markers revealed that both treatments led to a reduction in the LC3-II/LC3-I ratio and an accumulation of p62 protein, indicative of impaired autophagic flux (Figure 6B). Given that PKD2 is a direct target of miR-17-5p, we next assessed whether modulation of PKD2 expression mediates the effects of miR-17-5p on autophagy. As shown in Figure 6C, transfection with the miR-17-5p mimics suppressed autophagy, as reflected by a decreased LC3-II/LC3-I ratio and increased p62 levels. Notably, co-transfection with a PKD2 overexpression plasmid partially rescued these autophagic defects. Collectively, these findings demonstrate that PKD2 promotes basal autophagic activity by regulating intracellular Ca^{2 +} dynamics, and that miR-17-5p may impair autophagy through its inhibitory effect on PKD2.

PKD2 promotes mitochondrial autophagy and facilitates CSFV replication via the AMPK/mTOR signaling pathway

Building upon the observation that the PKD2-Ca^{2 +} axis regulates autophagy and considering the pivotal role of the AMPK/mTOR pathway in autophagic regulation [38,39], we further explored whether PKD2 modulates CSFV replication through this pathway.PK-15 cells were infected with CSFV at an MOI of 1. Western blot analysis showed that (Figure 7A and 7) the ratio of the autophagy marker LC3-II to LC3-I was significantly increased (p < 0.0001), while p62 protein levels were significantly diminished (p < 0.01) compared to the control group. These results indicate that CSFV infection activates autophagic flux and accelerates degradation of the autophagic substrate p62.Western blot analysis following PKD2 silencing demonstrated (Figure 7C and 7) significantly elevated mTOR phosphorylation (p-mTOR/mTOR) compared to control group (p < 0.05), whereas AMPK phosphorylation (p-AMPK/AMPK) was markedly reduced (p < 0.01). These results indicate that PKD2 deficiency suppresses AMPK activation, which consequently attenuates the AMPK-mediated inhibition of mTOR.PK-15 cells were subsequently treated with an AMPK inhibitor (Compound C, 10 μM) or an mTOR inhibitor (Sapanisertib, 100 nM), followed by CSFV infection at MOI = 1. Western blot analysis revealed that in the AMPK-inhibited group (Compound C), phosphorylation of AMPK (Thr172) was significantly decreased (p < 0.05), while phosphorylation of mTOR (Ser2448) was increased (p < 0.05). These changes were associated with a reduced LC3-II/LC3-I ratio (p < 0.01), increased p62 protein accumulation (p < 0.05), and a significant decrease in CSFV E2 protein expression (p < 0.01)(Figure 7E–7), Combined with the TCID₅₀ method to determine virus titers, analysis showed that in the AMPK-inhibited group (Compound C) exhibited a significant reduction in viral titer (p < 0.0001, Figure 7I), while in the mTOR inhibitor group (Sapanisertib) exhibited a significant increased in viral titer (p < 0.0001, Figure 7I), indicating suppressed autophagic activity and viral replication. Conversely, inhibition of mTOR with Sapanisertib led to increased AMPK phosphorylation (p < 0.05) and decreased mTOR phosphorylation (p < 0.05), alongside an elevated LC3-II/LC3-I ratio (p < 0.01), reduced p62 level (p < 0.05), and significantly enhanced CSFV E2 protein expression (p < 0.001), reflecting enhanced autophagic flux and viral replication. Interestingly, in the group treated with both Compound C and Sapanisertib, none of these parameters showed significant changes compared to controls, suggesting that the pro-autophagic and proviral effects of PKD2 are dependent on the functional integrity of the AMPK/mTOR pathway.Collectively, these findings demonstrate that PKD2 promotes autophagic flux by activating the AMPK pathway and suppressing mTOR signaling, thereby creating a cellular environment conducive to CSFV replication and facilitating viral proliferation.

Figure 7.

PKD2 regulates autophagic flux via the AMPK/mTOR pathway to promote CSFV proliferation. (A) PK-15 cells were infected with CSFV (MOI = 1), and the expression of autophagy-related markers (LC3I, LC3II, and p62) and CSFV E2 protein was quantified by Western blotting at 48 hours post-infection. (B) quantitative analysis of band intensity was performed using image J software to evaluate autophagy changes induced by CSFV infection. (C) PK-15 cells were transfected with siPKD2-2 for 12 hours. Western blotting was performed to detect the expression of total AMPK, phosphorylated AMPK (Thr172), total mTOR, and phosphorylated mTOR (Ser2448). (D) densitometric quantification of protein bands was carried out using image J software to assess alterations in pathway activity after PKD2 silencing. (E) PK-15 cells were treated with either an AMPK inhibitor (Compound C, 10 μM, 6 hours) or an mTOR inhibitor (Sapanisertib, 100 nM, 6 hours). Cells were then infected with CSFV (MOI = 1), and samples were harvested 24 h post-infection for protein analysis. Western blotting was performed to detect autophagy-related markers (LC3I, LC3II, and p62), key components of the AMPK/mTOR signaling pathway (total and phosphorylated AMPK [Thr172], total and phosphorylated mTOR [Ser2448]), and CSFV E2 protein expression. (F-H) quantitative densitometric analysis of protein bands was conducted using image J software to evaluate alterations in pathway activity and autophagic flux. (I) for viral titer analysis, PK-15 cells were treated with either an AMPK inhibitor (Compound C, 10 μM, 6 hours) or an mTOR inhibitor (Sapanisertib, 100 nM, 6 hours), followed by CSFV inoculation using ten-fold serial dilutions. After 3 days, viral titers were determined using the TCID₅₀ assay and expressed as Log₁₀ TCID₅₀/0.1mL. Data are presented as mean ± standard deviation (SD), with error bars representing SD from three independent biological replicates. Statistical significance was determined using Student’s t-test and one-way ANOVA. Asterisks denote significant differences compared to control: ns, not significant; *p<0.05; **p < 0.01; ***p < 0.001. ****p < 0.0001. Representative blots are shown from three independent experiments.

Discussion

MicroRNAs (miRNAs), as an important class of endogenous small non-coding RNAs, play crucial roles in antiviral defense mechanisms [40]. The miR-17 ~ 92 cluster, in particular, has been implicated in the regulation of viral replication and host immune responses. For instance, miR-92a, a member of this cluster, suppresses mouse mammary tumor virus (MMTV) replication [41], while enterovirus 71 inhibits its own replication by enhancing promoter methylation of the miR-17 ~ 92 cluster [42]. Furthermore, miR-17–92 is essential for the activation of CD4 Th1 and T follicular helper cell during viral infections [43]. Among its members, miR-17-5p has emerged as a key regulator of viral replication via diverse molecular pathways [12–15].

Current research on porcine viruses and autophagy pathways indicates that the regulation of autophagy by viruses is primarily mediated by calcium ions. For instance, the porcine reproductive and respiratory syndrome virus (PRRSV) activates CaMKII-dependent autophagy by inducing a Ca^{2 +} imbalance [22]. Similarly, porcine circovirus type 2 (PCV2) utilizes the CaMKKβ-mediated autophagy pathway, which can be inhibited by Ca^{2 +} signaling [23]. Additionally, rotavirus (RV) employs the viral channel protein NSP4 to deplete the endoplasmic reticulum calcium pool, thereby triggering an endoplasmic reticulum stress autophagy cascade [24]. These findings suggest that calcium ion-mediated autophagy regulation represents a critical component in the interaction between viruses and their porcine hosts. Nevertheless, there remains a relative paucity of research concerning the relationship between viral autophagy and calcium ions in classical swine fever virus (CSFV) and its autophagy regulation.In this study, we elucidate for the first time the molecular mechanism by which classical swine fever virus (CSFV) infection restricts its own replication via upregulation of the host miR-17-5p. Specifically, miR-17-5p directly targets and suppresses the expression of polycystin-2 (PKD2), a calcium channel protein, thereby inhibiting AMPK/mTOR pathway-mediated autophagic flux. Experimental evidence demonstrated that CSFV infection significantly induced miR-17-5p expression in PK-15 cells (Figure 1A). Overexpression of miR-17-5p via mimics transfection resulted in a dose-dependent reduction of CSFV E2 gene and protein levels (Figure 2A-C), as well as decreased viral titers (Figure 2G). This antiviral effect was mechanistically linked to PKD2 inhibition, as siRNA-mediated PKD2 knockdown similarly inhibited viral replication (Figure 5), diminished intracellular Ca^{2 +} release (Figure 6A), and impaired autophagic flux – indicated by a decreased LC3-II/LC3-I ratio and p62 accumulation, Fig. 6B). Simultaneously based on the results shown in Figures 6A and 7C-D, it is reasonable to speculate that PKD2 May influence the AMPK/mTOR pathway by modulating intracellular calcium homeostasis, thereby affecting autophagy and viral replication.These findings challenge the prevailing paradigm that members of the Flaviviridae family universally exploit miR-17-5p to enhance viral replication, instead revealing a host defense function of miR-17-5p in the context of CSFV infection. Interestingly, bovine viral diarrhea virus (BVDV), another Pestivirus genus member, employs an opposing mechanism: BVDV requires miR-17-5p binding to its 3’-UTR S2 site to initiate viral translation [14]. In contrast, CSFV-induced miR-17-5p expression functions to restrict viral replication by downregulating PKD2. This mechanistic divergence may reflect differences in viral replication strategies, with BVDV assembling at cytoplasmic membrane structures that are less dependent on autophagy [44], whereas CSFV exploits autophagosomal membranes as replication scaffolds.

PKD2 functions as a critical regulatory node linking Ca^{2 +} signaling to autophagic pathways [29]. In the context of CSFV infection, there are three distinct subcellular localization patterns through which PKD2 mediates its effects: (1) at the plasma membrane, where it forms a functional complex with PKD1 to facilitate calcium influx [23,45]; (2) on endoplasmic reticulum membranes, where it interacts with IP3R to enhance calcium release from intracellular stores [46]; and (3) on ciliary membranes, where its channel activity is inhibited by TACAN [30]. Both PKD2 knockdown and calcium chelation using BAPTA-AM significantly reduced the LC3-II/LC3-I ratio and led to p62 accumulation (Figure 6B), confirming that PKD2 positively regulates basal autophagy.

Notably, autophagy exerts stage-dependent effects on CSFV replication. During the early phase of infection, autophagy may enhance antiviral immunity by promoting the presentation of viral antigen. In contrast, at later stages, autophagy is often subverted by the virus to generate membranous scaffolds that facilitate its replication [19]. The proviral effect of the PKD2- mediated autophagy axis observed in our study (Figure 5 and 7) is consistent with previous findings that the CSFV NS5A protein-induced autophagy to support viral propagation [19,47]. By comparison, other viruses such as poliovirus evades host degradation mechanisms by blocking autophagosome maturation [48], highlighting the diverse, viruse-specific strategies used to exploit or evade autophagic processes.

Recent studies have demonstrated that African swine fever virus (ASFV) replicates exclusively in the cytoplasm [49], with viral DNA replication and transcription confined to perinuclear viral factories (VFs) [50] and independent of nuclear involvement [49]. The ASFV-encoded E301R protein assembles into a tetrameric sliding clamp structure and directly interacts with the viral DNA polymerase (O174L), functioning analogously to host proliferating cell nuclear antigen (PCNA) [51]. Unlike ASFV’s strategy of directly encoding replication machinery, our study reveals that CSFV, an RNA virus, exploits the host PKD2-autophagy axis to facilitate replication. This highlights a fundamental divergence in replication strategies between DNA and RNA virus, wherein the latter often co-opt host signaling pathways rather than encoding analogous machinery.

Interestingly, miR-17-5p has been reported to promote lung adenocarcinoma progression by targeting PKD2 [52]. However, in our CSFV infection model, miR-17-5p exerts an antiviral effect through PKD2 suppression, thereby inhibiting autophagy hijacked by the virus.This apparent functional dichotomy likely reflects distinct microenvironmental contexts between tumor progression and viral infection. In the setting of infection, PKD2-mediated autophagy is frequently hijacked by viruses to support replication. For example, poliovirus utilizes autophagosomes as replication- competent structures [53], coxsackievirus B3 (CVB3), actively induces autophagosome formation [54], and measles virus maintains autophagic signaling to enhance intercellular spread [55]. Consequently, PKD2 inhibition can counteract viral replication by disrupting this autophagic advantage.

Given the crucial role of the calcium-autophagy signaling axis in supporting viral life cycles [26], as observed in influenza A virus (IAV), which reduces intracellular calcium via L-type calcium channel (LTCC/Cav1.3) inhibition to promote autophagosome accumulation [56], porcine reproductive and respiratory syndrome virus (PRRSV), which activates autophagy through CaMKII [24], and porcine circovirus type 2 (PCV2), which depends on CaMKKβ pathway [25], our data suggest that PKD2 may represent a promising broad-spectrum antiviral target, particularly in porcine viral infections.

Building upon previous studies [57,58], we propose a mechanistic model to illustrate the regulatory axis between host miR-17-5p and CSFV replication (Figure 8): First, CSFV infection induces the upregulation of miR-17-5p in host cells. Second, PKD2—identified as a direct target of miR-17-5p—ordinarily facilitates CSFV proliferation by promoting autophagic flux through activation of the AMPK/mTOR signaling pathway. Third, as an RNA virus replicating exclusively in the cytoplasm, CSFV replication depends on PKD2-mediated Ca^{2 +} signaling to sustain autophagy, particularly through endoplasmic reticulum-localized PKD2, which enhances IP3R-dependent calcium release [59,60]. Finally, the CSFV-triggered increase in miR-17-5p suppresses PKD2 translation, thereby disrupting autophagy and ultimately restricting CSFV replication.

Figure 8.

Schematic model illustrating the interaction between CSFV and host miR-17-5p. In PK-15 cells, CSFV infection induces upregulation of host miR-17-5p expression. Under normal conditions, PKD2 promotes autophagic flux via activation of the AMPK/mTOR signaling pathway, thereby facilitating CSFV replication. However, miR-17-5p directly targets the 3’-UTR of PKD2 mRNA, leading to the suppression of PKD2 expression. This miR-17-5p-mediated downregulation of PKD2 attenuates autophagy, thereby limiting CSFV proliferation. These findings suggest a host defense mechanism in which miR-17-5p negatively regulates CSFV replication by targeting the pro-autophagic factor PKD2.

To further substantiate these findings, future investigations should pursue the following directions: (1) employ super-resolution confocal imaging (STED) and co-immunoprecipitation (Co-IP) assays to elucidate the subcellular localization and interaction partners of PKD2 during CSFV infection; (2) explore whether CSFV nonstructural proteins – particularly NS5A – directly regulate PKD2 activity or modulate miR-17-5p expression; (3) conduct comprehensive analyses of miR-17-5p and PKD2 expression profiles across various CSFV-infected tissues; (4) assess the spatiotemporal dynamics of miR-17-5p–PKD2 interactions and their functional implications in regulating CSFV infection in vivo; and (5) in-depth investigation into PKD2-mediated regulation of intracellular calcium dynamics and the causal signaling cascade underlying AMPK/mTOR pathway activation.

In conclusion, based on experimental research and plausible deductions from Ca²⁺ - AMPK/mTOR literature [61,62], this study systematically delineates the miR-17-5p-PKD2-Ca^{2 +} -AMPK/mTOR-autophagy signaling cascade in the context of CSFV infection (Figure 8), and a new defense mechanism is suggested for hosts to combat virus replication through the regulation of the “miRNA ion channel autophagy” axis. By demonstrating that CSFV-induced miR-17-5p restricts viral replication through targeted suppression of PKD2 and disruption of virus-exploited autophagic flux, our findings provide critical insight into the host-pathogen interaction landscape. Although our study does not directly establish the causal link between calcium dynamics and AMPK/mTOR pathway. Future work will be directed toward experimentally validating this regulatory axis to address this important gap. This mechanistic framework not only advance our understanding of Pestivirus pathogenesis but also offers a conceptual basis for the development of host-targeted antiviral strategies against CSFV.

Funding Statement

This work was supported by the Yunnan Revitalization Talent Support Program Team (Project No.: 202405AS350004) and the Scientific Research Fund Project of the Yunnan Department of Education (Project No.: 2025Y0477). The funder were not involved in the study design, data collection and analysis, or the decision to submit the manuscript for publication.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The source data of this study are available in figshare with doi number (https://doi.org/10.6084/m9.figshare.30847397) [63].