TRIM29 knockout pigs exhibit enhanced broad-spectrum disease resilience by amplifying type I interferon antiviral defenses

Abstract

While production traits have seen accelerated genetic improvement through advanced breeding technologies, disease resilience phenotypes continue to pose significant challenges in livestock breeding system. Current gene editing technologies provide an effective and biosafe strategy to enhance livestock disease resilience through precise manipulation of host antiviral genes. In this study, we successfully generated disease-resilient pigs exhibiting broad-spectrum antiviral activity against multiple viruses, with no observed adverse effects on pig health. The E3 ubiquitin ligase TRIM29 functions as a negative regulator of type I interferon (IFN) signaling, thus representing a potential antiviral target. Knockdown of TRIM29 in PK15 cells significantly enhanced antiviral immunity against pseudorabies virus (PRV) and vesicular stomatitis virus (VSV) by augmenting type I IFN production. Translationally, we generated Trim29 knockout (KO) mice and confirmed their enhanced antiviral ability to both PRV and VSV infections. Subsequently, we produced TRIM29-KO pigs via gene editing coupled with somatic cell nuclear transfer. Compared to wild-type controls, the TRIM29-KO pigs exhibited significantly enhanced resilience to PRV infection, which was associated with elevated type I IFN levels in vivo. Furthermore, alveolar macrophages derived from TRIM29-KO pigs showed reduced susceptibility to infection with PRV, VSV, and transmissible gastroenteritis virus (TGEV), highlighting their potential broad-spectrum antiviral activity against multiple viral pathogens.

Author summary

Pigs were domesticated about 9000 years ago, but infectious diseases still represent the biggest constraint on modern pig production. Traditional breeding has improved growth and lean meat content, yet cannot keep up with rapidly evolving viral pathogens. The 20th and 21st centuries have seen severe pig pandemics that caused massive global economic losses. Conventional disease control, relying on biosecurity, vaccination, and eradication, is inherently reactive, often entailing high costs and suboptimal efficacy. The disease burden urgently calls for a shift from disease control to genetic resilience in pigs. Gene editing is a practical tool for agricultural genetic improvement. Several edited crops and livestock have gained commercial approval recently, making it feasible to breed disease-resilient pigs and break the epidemiological bottleneck. We here generated the gene-edited pigs with enhanced disease resilience, by knockout of the negative regulator of type I interferon (IFN), TRIM29. Upon PRV infection, these pigs showed improved survival and milder symptoms, correlating with stronger antiviral IFN response. Primary cells from the edited pigs were also resilient to VSV and TGEV infection. This work provides a novel pig resource with broad-spectrum disease resilience, offering potential benefits for improving pig health and combating pandemics in the pig industry.

Introduction

Current intensive farming practices have significantly boosted livestock production by integrating controlled environments, robust biosecurity, and precision nutrition. Nevertheless, these same conditions have fostered the widespread transmission of infectious diseases, which poses a major threat to the economic sustainability of global pork production. In addition to conventional approaches such as vaccines and pharmaceuticals, genetic breeding strategies aimed at enhancing disease resilience in pigs present a promising alternative for effective disease control. By identifying genetic loci associated with specific diseases, gene editing technologies allow scientists to make precise genetic modifications to the pig genome, thereby creating heritable disease-resilient pig breeds. This approach has already demonstrated success in developing pigs resistant to porcine reproductive and respiratory syndrome virus (PRRSV). This was achieved by knocking out (KO) or precisely modifying the virus-interacting domain of the viral receptor gene CD163, a genetic alteration that effectively blocks PRRSV infection [1–3]. Subsequently, the same strategy has been applied to confer resistance against other pathogens. For example, DNAJC14 KO or a specific single-amino-acid substitution, as well as APN KO, have been shown to provide complete resistance to classical swine fever virus (CSFV) and transmissible gastroenteritis virus (TGEV) infection, respectively [4,5]. By offering a safe and ethically acceptable approach, gene editing via intentional genomic alteration (IGA) represents a promising biotechnology for improving desirable traits in livestock breeding and has achieved regulatory approval for commercialization in several countries [6].

Pathogen resilience in animals arises from a multi-layered defense system that integrates innate immunity, adaptive immune memory, and targeted molecular countermeasures—each fine-tuned through evolutionary host-pathogen dynamics—to block infection and ensure survival. Type I interferon (IFN) is the critical mediator of innate immunity, acting as the frontline sentinel against viral pathogens in animals. Type I IFN defends against viruses through induction of various antiviral effectors encoded by IFN-stimulated genes (ISG) [7–9]. Higher IFN level or ISG level are generally indicative of enhanced antiviral capacity at both cellular and organismal levels. The therapeutic potential of type I IFN has been widely utilized in clinical settings for managing diverse viral infections [10–13]. Notably, transgenic murine models engineered to constitutively overexpress type I IFN or specific ISG have demonstrated significant resilience to viral challenges in numerous experimental studies [14,15]. These findings collectively underscore the critical role of the type I IFN response in establishing antiviral defenses in animals. While the type I IFN response plays a crucial protective role in antiviral defense, systemic IFN therapy or its constitutive overexpression can induce detrimental consequences through excessive inflammation and tissue damage [9,16,17]. To maintain homeostasis, the host has evolved sophisticated regulatory mechanisms that ensure precise spatiotemporal control of type I IFN production. This fine-tuned balance is achieved through the coordinated interplay of multiple positive and negative factors.

TRIM29 serves as a negative regulator of type I IFN signaling, functioning through its canonical role as an E3 ubiquitin ligase. By ubiquitinating key signaling components in the type I IFN pathway, TRIM29 targets them for proteasomal degradation, thereby suppressing downstream IFN production [18]. This activity establishes TRIM29 as a dedicated negative feedback regulator of the production of type I IFN. During DNA virus infection, TRIM29 induces ubiquitination and proteolytic degradation of STING, thereby limiting type I IFN production [19,20]. In RNA virus infection, TRIM29 has been shown to ubiquitinate and degrade MAVS or NEMO to inhibit the type I IFN signaling pathway [18,21]. Beyond its role in modulating type I IFN, TRIM29 also suppresses the production of type III interferon (IFN-λ) and interleukin-18 (IL-18) by intestinal epithelial cells during enteric RNA virus infection. This suppression is achieved through the ubiquitination and degradation of NLRP6 and NLRP9b [22]. The ability of TRIM29 to target a diverse array of substrates underscores its broad involvement in the pathogenesis of both DNA and RNA viruses. In this context, TRIM29 deficiency enhances resilience against a variety of viral pathogens, including influenza virus, reovirus, Epstein-Barr virus, herpes simplex virus, rotavirus, and encephalomyocarditis virus, as demonstrated across multiple cellular systems and murine models [18–23].

Building on previous studies, TRIM29 has been identified as a promising candidate gene for breeding disease-resilient livestock with minimal adverse effects. Research in Trim29-KO mouse models has shown resilience to a broad spectrum of viral pathogens, while exhibiting no discernible phenotypic abnormalities beyond their enhanced antiviral capacity. Leveraging these insights, the current study aims to develop TRIM29-KO pigs and assess their antiviral potential, with the goal of cultivation of novel pig breed with enhanced disease resilience to benefit pig production.

Results

TRIM29 regulates PRV infection in PK15 cells

We first used pseudorabies virus (PRV) to investigate how TRIM29 regulates DNA virus infection in pig cells. Quantitative PCR (qPCR) and immunoblot results showed that PRV infection in PK15 cells induced TRIM29 expression in both mRNA and protein levels (Fig 1A-1C). PRV infection also induced a significant elevation of IFNβ mRNA in PK15 (Fig 1D). We designed 3 siRNAs targeting pig TRIM29 and tested their inhibitory effect on TRIM29 expression in mRNA and protein levels in PK15 (Fig 1E-1G). The siRNA targeting at TRIM29 mRNA locus starting from 882 bp (si882) was used in following gene silencing experiments for its stable inhibitory effect on TRIM29 expression. To investigate how TRIM29 regulates IFNβ expression and affects PRV infection, PK15 cells were transfected with si882 or negative control siRNA (siNC) for 24 h prior to PRV infection. At 24 h post-infection, IFNβ expression was significantly higher in si882-treated cells than in siNC controls, indicating that TRIM29 knockdown potentiates IFNβ induction (Fig 1H). Consistent with this, PRV DNA copies and viral titers in the culture supernatant were markedly lower in the si882 group compared to the siNC group, correlating with elevated IFNβ levels (Fig 1I and 1J). As a positive control, cGAMP treatment similarly induced higher IFNβ expression in si882-transfected cells (Fig 1H). To further validate TRIM29 modulating effect on IFNβ expression, we transiently transfected a plasmid overexpressing the pig TRIM29 with the blank plasmid vector as the control (Fig 1K and 1L). TRIM29 overexpression did not alter constitutive IFNβ expression in uninfected PK15 cells but significantly inhibited cGAMP- and PRV-induced IFNβ transcription (Fig 1M), thereby promoting viral replication in TRIM29-overexpressing PK15 cells (Fig 1N and 1O). We next infected PK15 cells with PRV at both low (0.1) and high (1) multiplicity of infection (MOI) and monitored viral replication kinetics at multiple time points post-infection. PRV replication was consistently lower in TRIM29-knockdown cells and higher in TRIM29-overexpressing cells compared to controls, confirming that TRIM29 positively regulates PRV infection in PK15 cells (Fig 1P-1S).

Fig 1. TRIM29 supports PRV infections by inhibiting type I IFN production.

A. Total RNA was isolated from PK15 cells infected with PRV at MOI = 0.1 for 24 h and the relative mRNA expression of TRIM29 was detected by qPCR with GAPDH as the internal gene control. B. Representative immunoblot analysis of TRIM29 and β-actin in PK15 cells infected with PRV at MOI = 0.1 for 24 h. C. Quantification of protein levels by analyzing band density shown in B using Image J software. D. The relative mRNA expression of IFNβ was detected by qPCR with GAPDH as the internal gene control in PK15 cells infected with PRV at MOI = 0.1 for 24 h. E. Three pig TRIM29-specific siRNA were design for gene silencing in PK15 cells. The TRIM29 mRNA levels were analyzed by qPCR in PK15 cells transfected with the indicated siRNA for 36 h. F. Representative immunoblot analysis of TRIM29 protein in siRNA-transfected PK15 cells. G. Quantification of TRIM29 protein levels shown in F by analyzing band density. H. qPCR of IFNβ mRNA in PK15 cells transfected with si882 and control siNC for 24 h and then infected with PRV at MOI = 0.1 or treated with cGAMP for 24 h. I. PRV DNA genome copies were determined in culture supernatant of PK15 cells transfected with si882 and NC siRNA for 24 h and then infected with PRV at MOI = 0.1 for 24 h. J. PRV viral titers determined by TCID₅₀ in the TRIM29-knockdown PK15 with PRV infection. K. Representative immunoblot of TRIM29 protein in PK15 cells transfected with TRIM29-overexpressing plasmid and blank vector as the control. L. Quantification of overexpressing TRIM29 by measuring band density. M. The TRIM29-overexpressing PK15 were infected with PRV at MOI = 0.1 or treated with cGAMP for 24 h and the relative mRNA expression of IFNβ was detected by qPCR. N. PRV DNA copies were determined in the TRIM29-overexpressing PK15 with PRV infection. O. PRV viral titers determined by TCID₅₀ in the TRIM29-overexpressing PK15 with PRV infection. P. PRV growth curve in TRIM29-knockdown and control PK15 cells inoculated at MOI = 0.1. Q. PRV growth curve in TRIM29-knockdown and control PK15 cells inoculated at MOI = 1. R. PRV growth curve in TRIM29-overexpressing and control PK15 cells inoculated at MOI = 0.1. S. PRV growth curve in TRIM29-overexpressing and control PK15 cells inoculated at MOI = 1. T. Immunoblot analysis of TRIM29, cGAS, STING, TBK1, phosphorylated TBK1 (pTBK1), IRF3, phosphorylated IRF3 (pIRF3), and GAPDH in PK15 cells transfected with TRIM29 siRNA (si882, left panel) and overexpressing vector (right panel) with mock and PRV infection. Data are expressed as mean ± SD. The p-values between the indicated groups were calculated using a two-tailed unpaired t-test (A, C, D, H-J, L and N-S) or one-way ANOVA with Sidak’s multiple comparison test (E and G) to determine statistical significance.

To investigate the molecular mechanism underlying TRIM29-mediated regulation of type I IFN production, we examined the key components of the cGAS-STING signaling pathway, which governs DNA virus-induced type I IFN responses. Immunoblot analysis revealed markedly increased protein levels of STING, phosphorylated TBK1 (pTBK1), and phosphorylated IRF3 (pIRF3) in TRIM29-knockdown cells following PRV infection (Fig 1T, left panel), suggesting that TRIM29 suppresses STING stabilization and subsequent type I IFN production. Conversely, TRIM29 overexpression significantly reduced STING expression which sensing PRV infection, consequently impairing TBK1-IRF3 signaling activation (Fig 1T, right panel). These findings establish TRIM29 as a negative regulator of the cGAS-STING axis, wherein it degrades STING to constrain DNA virus-triggered IFN responses.

TRIM29 regulates VSV infection in PK15 cells

Given that TRIM29 can target protein substrates in the type I IFN pathways mediated by both DNA and RNA viruses, we next investigated whether TRIM29 modulates RNA virus infection in pig cells using vesicular stomatitis virus (VSV) as a model. Unlike PRV, VSV infection downregulated TRIM29 mRNA levels and induced more than 150-fold higher expression of IFNβ in PK15 cells (Fig 2A and 2B). siRNA-mediated TRIM29 knockdown significantly enhanced VSV-induced IFNβ mRNA expression while having no effect on basal IFNβ levels (Fig 2C). This enhanced antiviral response correlated with reduced VSV replication, as evidenced by decreased viral copies and titers in culture supernatant (Fig 2D and 2E) and lower VSV positivity in PK15 cells (Fig 2I). Conversely, TRIM29 overexpression attenuated VSV-induced IFNβ expression and amplified VSV replication in PK15 cells (Fig 2F-2H and 2J). Analysis of VSV replication kinetics over time showed that TRIM29 knockdown impaired viral growth in PK15 cells (Fig 2K and 2L), while its overexpression promoted replication (Fig 2M and 2N). Immunoblot analysis revealed no detectable alterations in MAVS protein levels in either TRIM29-knockdown or TRIM29-overexpressing PK15 cells following VSV infection, suggesting that MAVS is unlikely to be a downstream effector of TRIM29 in modulating VSV infectivity in PK15 cells. Notably, TRIM29-mediated suppression of STING protein was found under VSV infection conditions, implying that TRIM29 specifically regulates STING expression to affect TBK1-IRF3 axis activation during VSV infection in PK15 cells (Fig 2O).

Fig 2. TRIM29 facilitates VSV infections by inhibiting type I IFN production.

A. The relative mRNA expression of TRIM29 was determined by qPCR in PK15 cells with mock and VSV infection at MOI = 0.1 for 24 h. B. The relative mRNA expression of IFNβ was determined by qPCR in PK15 cells with mock and VSV infection at MOI = 0.1 for 24 h. C. Determination of IFNβ level in TRIM29-knockdown PK15 cells infected with VSV at MOI = 0.1 or treated with poly (I:C) for 24 h. D. Determination of VSV RNA copies in the supernatant of TRIM29-knockdown cells infected with VSV at MOI = 0.1 for 24 h. E. Determination of VSV titers in TRIM29-knockdown cells infected with VSV at MOI = 0.1 for 24 h. F. The relative mRNA expression of IFNβ was determined by qPCR in TRIM29-overexpressing PK15 cells with mock and VSV infection at MOI = 0.1 for 24 h or treated with poly (I:C) for 24 h. G. Determination of VSV RNA copies in the supernatant of TRIM29-overexpressing cells. H. Determination of VSV titers in the supernatant of TRIM29-overexpressing cells. I. EGFP fluorescence showing VSV infection level in TRIM29-knockdown PK15 cells infected with VSV at MOI = 0.1 for 24 h. J. EGFP fluorescence in TRIM29-overexpressing PK15 cells infected with VSV at MOI = 0.1 for 24 h. Mock represents uninfected groups. Scale bars, 100 μm. K. VSV growth curve in TRIM29-knockdown and control PK15 cells inoculated at MOI = 0.1. L. VSV growth curve in TRIM29-knockdown and control PK15 cells inoculated at MOI = 1. M. VSV growth curve in TRIM29-overexpressing and control PK15 cells inoculated at MOI = 0.1. N. VSV growth curve in TRIM29-overexpressing and control PK15 cells inoculated at MOI = 1. O. Immunoblot analysis of TRIM29, MAVS, STING, TBK1, pTBK1, IRF3, pIRF3, and GAPDH in PK15 cells transfected with si882 (left panel) and TRIM29-overexpressing vector (right panel) with mock and VSV infection. Data are expressed as mean ± SD. The p-values between the indicated groups were calculated using a two-tailed unpaired t-test (A-H and K-N).

Trim29-KO mice are resilient to PRV and VSV infection

To investigate the antiviral role of TRIM29 in vivo, we generated Trim29-KO mice via CRISPR/Cas9-mediated gene editing. Successful disruption of Trim29 gene and the absence of TRIM29 protein in KO mice were confirmed by PCR genotyping and immunoblot analysis, respectively (Fig 3A and 3B). Growth curve analysis revealed that homozygous Trim29-KO mice exhibited significantly greater weight gain during the pre-weaning period compared to heterozygous KO and wild-type (WT) littermates, with this trend observed in both male and female pups. Notably, post-weaning growth rates were comparable across all three genotypes (Fig 3C and 3D).

Fig 3. Establishment of Trim29-KO mice and assessment of their antiviral ability.

A. Genotyping of Trim29 homozygous (homo) KO, heterozygous (hetero) KO, and WT mice. Blank, no template controls in PCR reaction. Ladder, DNA ladders. B. Representative immunoblot of TRIM29 protein levels in skin tissues of homo KO, hetero KO, and WT mice. C. Body weight curves of male homo KO, hetero KO, and WT mice. D. Body weight curves of female homo KO, hetero KO, and WT mice. E. The differences in phenotypic manifestations in the three genotypes of mice infected with PRV. F. Survival curves of the three genotypes of mice infected with PRV. G. Relative IFNα mRNA expression in brains of mice infected with PRV. H. Relative IFNβ mRNA expression in brains of mice infected with PRV. I. PRV DNA load in brains of infected mice. J. PRV DNA load in lungs of infected mice. K. PRV DNA load in serum of infected mice. L. Survival curves of the three genotypes of mice infected with VSV. M. Relative IFNα mRNA expression in brains of mice infected with VSV. N. Relative IFNβ mRNA expression in brains of mice infected with VSV. O. VSV RNA load in brains of infected mice. P. VSV RNA load in serum of infected mice. Data are expressed as mean ± SD. The p-values in growth curves were calculated using two-way ANOVA with Sidak’s multiple comparison test (C and D). Survival curve between groups were analyzed using the Gehan-Breslow-Wilcoxon test (F and L). The comparisons of viral loads and IFNɑ/β expression were conducted with one way ANOVA followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (I-K, N and O) or Kruskal-Wallis test followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (G, H, M and P).

Age-matched Trim29-KO mice (comprising homozygous and heterozygous KO genotypes) and WT controls were subcutaneously challenged with PRV and monitored over a 96-h period. Homozygous KO mice demonstrated attenuated clinical manifestations and significantly reduced mortality compared to both heterozygous KO and WT groups (Fig 3E and 3F). Analysis of type I IFN levels in brain tissues revealed a selective elevation of IFNα and IFNβ mRNA in homozygous KO mice (Fig 3G and 3H). Consistent with these findings, qPCR demonstrated markedly reduced PRV genomic loads in the brains of homozygous KO mice relative to heterozygous and WT groups (Fig 3I). Further assessment of viral loads in other tissues identified diminished PRV burden in the lungs of homozygous KO mice (Fig 3J). While virus in serum also trended lower in homozygous KO mice, this reduction did not reach statistical significance (Fig 3K).

VSV challenge data corroborated the enhanced antiviral phenotype in homozygous Trim29-KO mice. Throughout the 12-d challenge period, homozygous Trim29-KO mice exhibited complete survival (0% mortality), whereas heterozygous Trim29-KO and WT controls displayed approximately 57% survival (Fig 3L). Homozygous KO mice demonstrated significant upregulation of IFNα and IFNβ mRNA in brains (Fig 3M and 3N), accompanied by markedly reduced VSV genomic copies in brains and serum compared to heterozygous KO and WT groups (Fig 3O and 3P).

Generation of TRIM29-KO pigs by gene editing and animal cloning

Building on the antiviral effects observed in TRIM29 knockdown cells and Trim29-KO mice, we next established TRIM29-KO pigs to investigate their disease resilience against multiple viruses. The KO pigs were generated by introducing biallelic insertion/deletion (indels) in exon 1 of the TRIM29 gene in fetal fibroblasts (PFF) using CRISPR/Cas9, followed by somatic cell nuclear transfer (SCNT) with the gene-edited fibroblasts as donors. A total of 10 surrogates underwent embryo transfer, resulting in 5 pregnancies and the birth of 31 cloned piglets (Fig 4A). Of these offspring, 26 were delivered healthy, and the surviving cloned pigs exhibited no apparent abnormalities in developmental progression or general health status during subsequent observation (Fig 4B).

Fig 4. Generation of TRIM29-KO pigs.

A. The summary of pig cloning. B. The pictures of the TRIM29-KO pigs as newborns and at 5 months old. C. Gene editing results of the TRIM29 gene in cloned pigs. The blue box highlights the gRNA target sequence, while the red dashed line and letter indicate base deletion and insertion, respectively. WT, wild-type TRIM29; Alleles1 and 2 are the gene-edited TRIM29 sequences in cloned pigs. D. Immunoblot analysis of TRIM29 protein level in the skin of KO and WT pigs.

Genotyping analysis confirmed that all cloned pigs carried the same biallelic indels as the donor fibroblasts, with one allele harboring a 1-bp insertion and the other a 1-bp deletion in exon 1 of the TRIM29 gene (Fig 4C). To validate the KO phenotype, we assessed TRIM29 protein expression in the cloned pigs. Given prior reports of high TRIM29 expression in skin tissue, we performed immunoblotting on ear-derived skin biopsies. TRIM29 protein was undetectable in all cloned pigs, whereas WT controls exhibited robust TRIM29 expression in skin samples (Fig 4D).

We further assessed the precision of gene editing in the cloned pigs to rule out potential perturbations caused by off-target effects. To predict potential off-target sites in the TRIM29-KO pigs, we employed Cas-Offinder and CRISPOR, permitting up to four mismatches (with a maximum of one bulge). A total of 934 off-target sites were identified (S1 Table). To detect mutations at these sites, we performed whole-genome sequencing (WGS) on the TRIM29-KO pig and aligned the data to the SSC11.1 reference genome. Candidate off-target sites displaying sequence variations relative to the reference were further validated via PCR amplification in both the TRIM29-KO pig and the donor cells used for gene editing. Subsequent Sanger sequencing of the amplified products confirmed no detectable off-target mutations between the TRIM29-KO pig and the donor cells.

TRIM29-KO pigs are resilient to PRV infection

To investigate the antiviral capacity of TRIM29-KO pigs, we conducted PRV infection challenge experiments comparing KO and WT animals. Six KO pigs and 15 breed-matched WT pigs at the age of 5 months were infected with PRV intranasally, with 5 mL of viral solution (1 × 10^6.9 TCID₅₀/100 μL) administered per nostril. Following viral challenge, clinical symptoms were monitored and rectal temperatures were measured every day (Fig 5A). Pigs in KO group exhibited milder disease symptoms compared to WT group, characterized by more activity levels, lower rectal temperatures, and reduced mortality rates (Fig 5A-5C). By the endpoint of infection (day 9 post-challenge), three out of fifteen WT pigs survived, though they displayed severe illness or moribund states. In contrast, three out of six surviving KO pigs appeared clinically normal (Fig 5A). Oral viral shedding was consistently lower in KO pigs compared to WT pigs across the infection period (Fig 5D). Viral loads in the brain and lung were quantified via qPCR, revealing significantly reduced PRV loads in KO pigs in both brain and lung (Fig 5E and 5F). Serum analysis showed lower PRV loads in KO pigs at day 7 compared to WT controls; however, this difference was not statistically significant by day 9 (Fig 5G). Further analysis detected PRV-specific antibodies in serum, demonstrating that both PRV gB and gE antibody titers were markedly reduced in KO pigs at days 7 and 9 post-infection, consistent with reduced disease severity in the KO group (Fig 5H and 5I).

Fig 5. PRV challenge outcomes in TRIM29-KO pigs.

A. Clinical sign of TRIM29-KO and WT pigs challenged with PRV. The viral lesion was analyzed and scored in accordance with the predefined assessment criteria. B. Survival curves of TRIM29-KO and WT pigs challenged with PRV. C. Rectal temperature changes following PRV challenge. D. Oral viral shedding level between TRIM29-KO and WT pigs challenged with PRV. E. PRV nucleic acid levels in the brains of PRV-infected pigs. F. PRV nucleic acid levels in the lungs of PRV-infected pigs. G. Serum PRV nucleic acid levels on days 7 and 9 following PRV challenge. H. PRV gB antibody level on days 0, 7 and 9 post-challenge. I. PRV gE antibody level on days 0, 7 and 9 post-challenge. Data are expressed as mean ± SD. The survival curves between TRIM29-KO and WT pigs were analyzed using the Gehan-Breslow-Wilcoxon test (B). The viral shedding level were analyzed using the two-way ANOVA followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (D). The p-values between the indicated groups were calculated using a two-tailed unpaired t-test or Mann-Whitney test, depending on data distribution normality (C and E-I).

Serum levels of IFNα and IFNβ in KO pigs remained persistently higher than in WT pigs at both 7 and 9 days following viral challenge (Fig 6A and 6B), whereas no significant differences were detected between the groups at baseline (day 0). Tissue mRNA analysis demonstrated a notable elevation of IFNα in the brain and lungs of KO pigs (Fig 6C), along with a significant upregulation of IFNβ in the brain (Fig 6D), compared to WT controls. The ISG genes (MX1 and ISG15) exhibited a tendency toward increased expression levels in the brains of KO pigs with PRV infection (Fig 6E).

Fig 6. Type I IFN signal and histopathological changes in PRV-challenged pigs.

A. Serum IFNα levels on days 7 and 9 following PRV challenge detected by ELISA. B. Serum IFNβ levels on days 7 and 9 following PRV challenge detected by ELISA. C. IFNα mRNA levels in the brains and lungs of PRV-infected pigs detected by qPCR. D. IFNβ mRNA levels in the brains and lungs of PRV-infected pigs detected by qPCR. E. The mRNA levels of MX1 and ISG15 in the brains of PRV-infected pigs detected by qPCR. F. HE staining of TRIM29-KO and WT pigs infected with PRV at 9 d post‑challenge. Scale bars: 100 μm. G. PRV antigen immunohistology of TRIM29-KO and WT pigs infected with PRV at 9 d post‑challenge. Scale bars: 50 μm. Data are expressed as mean ± SD. The p-values between the indicated groups were calculated using a two-tailed unpaired t-test or Mann-Whitney test, depending on data distribution normality (C and E-I).

Histological analysis revealed that the tissues of KO pigs exhibited significantly alleviated histological lesions. In the lungs of WT pigs, the alveolar walls were thickened with inflammatory cell infiltration. In the brain tissues of WT pigs, perivascular cuffing was observed, characterized by inflammatory cell infiltration around the blood vessels (Fig 6F). Immunohistochemical analysis demonstrated markedly stronger and more widespread PRV antigen staining signals in both brain and lung tissues of WT pigs compared to their KO counterparts, a pattern that aligned with the observed exacerbation of histopathological damage and higher viral burden in WT pigs (Fig 6G).

The integrated virological, morphological, and disease phenotypic data indicate that KO pigs elicit a heightened type I interferon response during PRV infection, which correlates with reduced viral infectivity and pathogenicity in vivo.

TRIM29-KO pig-derived PAMs are resilient to multiple viruses

Given the broad-spectrum antiviral properties of type I IFN, we expanded our investigation to assess the susceptibility to multiple viruses in primary isolated pulmonary alveolar macrophages (PAMs). Consistent with the in vivo challenge results, PRV infection in PAMs of TRIM29-KO pigs triggered elevated expression of IFNα and IFNβ (Fig 7A and 7B), accompanied by reduced viral titer (Fig 7C), compared to the infection in WT PAMs. This pattern extended to RNA viruses, including VSV (Fig 7D-7F) and TGEV (Fig 7G-7I), where elevated type I IFN mRNA levels correlated with reduced viral titer. Time-course analysis under various MOI conditions showed that PRV, VSV, and TGEV all exhibited reduced replication kinetics in KO PAMs (Fig 7J-7O). Together, these results indicate that PAMs derived from TRIM29-KO pigs display heightened resilience to diverse viral infections, a phenomenon driven by their augmented type I IFN signaling.

Fig 7. PAMs isolated from TRIM29-KO pigs were resilient to multiple viruses.

A. IFNα mRNA levels in TRIM29-KO and WT PAMs with mock and PRV infection at MOI = 0.1. B. IFNβ mRNA levels in TRIM29-KO and WT PAMs with mock and PRV infection at MOI = 0.1. C. PRV titers in TRIM29-KO and WT PAMs. D. IFNα mRNA levels in TRIM29-KO and WT PAMs with mock and VSV infection at MOI = 0.1. E. IFNβ mRNA levels in TRIM29-KO and WT PAMs with mock and VSV infection at MOI = 0.1. F. VSV titers in TRIM29-KO and WT PAMs. G. IFNα mRNA levels in TRIM29-KO and WT PAMs with mock and TGEV infection at MOI = 0.1. H. IFNβ mRNA levels in TRIM29-KO and WT PAMs with mock and TGEV infection at MOI = 0.1. I. TGEV titers in TRIM29-KO and WT PAMs. J. PRV growth kinetics in WT vs. KO PAMs infected at an MOI of 0.1. K. PRV growth kinetics in WT vs. KO PAMs infected at an MOI of 1. L. VSV growth kinetics in WT vs. KO PAMs infected at an MOI of 0.1. M. VSV growth kinetics in WT vs. KO PAMs infected at an MOI of 1. N. TGEV growth kinetics in WT vs. KO PAMs infected at an MOI of 0.1. O. TGEV growth kinetics in WT vs. KO PAMs infected at an MOI of 1. Data are expressed as mean ± SD. The mRNA levels of IFNα and IFNβ were analyzed with two-way ANOVA with Sidak’s multiple comparison test (A, B, D, E, G, and H). The viral copies and titers were analyzed with one-tailed unpaired t-test (C, F, and I) and two-tailed unpaired t-test (J-O).

TRIM29-KO mice and pigs do not exhibit elevated inflammation levels

Type I IFN modulates inflammatory responses, a critical biological process for defending against harmful stimuli and maintaining tissue homeostasis. Nevertheless, uncontrolled or dysregulated inflammatory responses can result in excessive tissue damage and drive chronic disease progression. Building on TRIM29's established role as a modulator of type I IFN signaling cascades, TRIM29-KO is likely to influence the production of both pro-inflammatory and anti-inflammatory cytokines upon specific pathophysiological contexts, which could constitute a non-negligible side effect of gene editing intervention. To compare the inflammation level and immune status between the KO and WT animals under homeostatic conditions, we quantified a series of pro- and anti-inflammatory cytokines in the serum. Furthermore, the immunoglobulins representing the immune status were quantitatively assessed. In the serum of nursery pigs (1 month), multiplex analysis demonstrated comparable concentrations of key cytokines, including IFNα, IFNγ, IL-10, IL-1β, IL-4, IL-6, TNFα, and IL-12, across genotypes. Notably, IL-8 exhibited significant downregulation in TRIM29-KO group compared to WT controls (Fig 8A). Furthermore, assessment of humoral immunity parameters showed no inter-group differences in immunoglobulin isotypes (IgG, IgA, IgM) as determined through ELISA assays (Fig 8B). In serum from adult (5-week-old) mouse models, pro-inflammatory factors TNFα, IFNγ, and IL17A exhibited reduced concentrations in Trim29-KO mice compared to controls, while concentrations of IL-1β, IL-6, and IL-10 remained unaltered (Fig 8C). Among immunoglobulin isotypes, serum IgA levels were reduced in KO mice, whereas IgG1, IgG2a, IgG2b, IgG3, IgE, and IgM showed no significant change (Fig 8D).

Fig 8. Comparative analysis of inflammatory and immunological profiles in TRIM29-KO versus WT pig and mouse models.

A. Luminex-based multiplex analysis of serum concentrations of 9 inflammatory cytokines in 1-month-old TRIM29-KO and WT pigs. B. ELISA quantification of serum levels of immunoglobulin isotypes (IgG, IgA and IgM) in 1-month-old TRIM29-KO and WT pigs. C. Luminex-based multiplex analysis of serum concentrations of 6 inflammatory cytokines in 5-month-old Trim29-KO and WT mice. D. Luminex-based multiplex analysis of serum concentrations of 7 immunoglobulin isotypes (IgG1, IgG2a, IgG2b, IgG3, IgA, IgE and IgM) in 5-month-old Trim29-KO and WT mice. E. Relative TNFα mRNA expression in KO vs WT PAMs following LPS and mock treatment. F. Relative IL-1β mRNA expression in KO vs WT PAMs following LPS and mock treatment. G. Relative IL-6 mRNA expression in KO vs WT PAMs following LPS and mock treatment. H. Relative COX2 mRNA expression in KO vs WT PAMs following LPS and mock treatment. I. Western blot analysis of pro-IL-1β, pro-TNFα, and COX2 protein levels between KO and WT PAMs after LPS and mock treatment. The p-values between the indicated groups were calculate using a two-tailed unpaired t-test or Mann-Whitney test, depending on data distribution normality (A-H).

To assess the inflammatory response under pathological conditions, we used lipopolysaccharide (LPS)-induced inflammation model in PAMs. PAMs derived from age- and breed-matched KO and WT pigs were treated with 20 ng/mL LPS for 12 h. The mRNA expression of pro-inflammatory cytokines (TNFα, IL-1β and IL-6) and the inflammation marker gene COX2 was then determined by qPCR. The results showed that the expression levels of these genes were comparable between KO and WT PAMs (Fig 8E-8H). Consistently, western blot analysis of cell lysates revealed similar protein levels of pro-TNFα, pro-IL-1β, and COX2 in both groups (Fig 8I).

Discussion

The TRIM family represents the largest subfamily of E3 ubiquitin ligases that target a wide range of protein substrates for ubiquitination and subsequent proteolytic degradation. TRIM proteins mediate ubiquitination of diverse regulatory components in type I IFN signaling pathways, thereby playing critical roles in modulating both the sensitivity and magnitude of IFN-mediated antiviral responses [24,25]. Accumulating evidence has established TRIM29 as a crucial negative regulator of type I IFN signaling through its E3 ubiquitin ligase activity targeting multiple pathway components [18–23]. Mechanistically, TRIM29 directly interacts with and ubiquitinates NF-κB essential modulator (NEMO), a pivotal signaling hub that integrates both IRF-mediated type I IFN production and NF-κB-dependent inflammatory responses, thereby suppressing influenza virus-induced IFN production in alveolar macrophages [18]. TRIM29 also modulates RNA virus sensing by targeting MAVS for K11-linked polyubiquitination and proteasomal degradation, effectively dampening reovirus-triggered type I IFN responses in bone marrow-derived dendritic cells [21]. In the context of DNA virus infection, TRIM29 exerts its immunosuppressive function by ubiquitinating STING, promoting viral replication of EBV and HSV-1 [19,20].

Building on TRIM29's well-characterized role as a stable negative regulator of type I IFN innate immunity through its targeting of multiple pathway components, Trim29-KO mice demonstrate markedly enhanced resilience to diverse viral pathogens. Crucially, these KO mice maintain normal growth, development, and fertility level, providing compelling evidence that TRIM29 ablation offers a promising strategy to achieve broad-spectrum antiviral protection in vivo without compromising essential physiological functions.

Our study begins by exploring the role of TRIM29 in pig cells using pig-tropic viruses known for their heightened sensitivity to type I IFN responses. Consistent with prior findings, siRNA-mediated TRIM29 silencing in PK15 cells markedly enhanced type I IFN production upon infection with PRV and VSV, leading to significant suppression of viral replication. Conversely, TRIM29 overexpression facilitated robust PRV and VSV proliferation by suppressing type I IFN induction. Mechanistically, we identified that TRIM29 targets the IRF3 signaling axis by destabilizing STING, a central adaptor in antiviral immunity. While STING is classically recognized as a DNA-sensing adaptor downstream of cGAS, our findings align with emerging evidence of a cGAS-independent STING pathway that orchestrates innate immune responses against enveloped RNA viruses [26,27]. This dual regulatory mechanism explains how TRIM29, through STING ubiquitination, modulates antiviral defenses against both DNA (PRV) and RNA (VSV) viruses in PK15 cells.

TRIM29-KO pigs were created in our work showed normal growth rate and physical development and the functional characterization of their antiviral capacity is the major research emphasis. As a type I IFN-sensitive virus, PRV is a major harmful pathogen and caused higher morbidity and mortality in pig populations. We therefore employed PRV as the challenge virus to evaluate the antiviral competence of TRIM29-KO pigs. PRV vaccination in surrogate sows leads to high-titer and long-lasting maternal antibody in the cloned pigs; the viral challenge experiments were thus performed at 5 months of age when the PRV gB antibodies were declined to the seronegative thresholds, as confirmed by both gB indirect ELISA and competitive ELISA methods. Following PRV challenge, WT pigs exhibited severe clinical outcomes, including high morbidity (15/15) and mortality (12/15). In contrast, TRIM29-KO pigs displayed significantly attenuated disease progression, with mortality rates reduced to 50% (3/6). Virological analysis demonstrated that KO pigs had significantly lower viral loads in key organs and serum antibody titers compared to WT controls. Additionally, KO pigs showed more production of IFNɑ and IFNβ, indicative of enhanced viral clearance mediated by higher IFNɑ and IFNβ. These findings collectively confirm that TRIM29 deficiency confers resilience to PRV infection by promoting type I IFN production and limiting viral replication. Besides increased type I IFN production, other untested factors here may also contribute to the enhanced antiviral capacity. TRIM29 has been reported to inhibit natural killer (NK) cell functions. TRIM29 deficiency leads to enhanced IFNγ production and antiviral NK cell activation [28]. Given the critical role of NK cells in innate immune defense against herpesviruses, including PRV, whether and how TRIM29 KO regulates NK cell activation in response to PRV infection warrants further investigation.

It is noteworthy that type I IFN serves as a broad-spectrum antiviral effector, suggesting that TRIM29-KO pigs may exhibit enhanced resilience to a diverse range of viruses. Using primary PAMs derived from TRIM29-KO pigs, we validated this hypothesis by demonstrating significantly improved resilience to PRV, VSV, and TGEV. However, the antiviral potency of TRIM29 deficiency is not universal to all viruses, as certain viruses employ distinct adaptors for viral sensing and IFN induction that may bypass TRIM29 regulation. Additionally, some viruses encode robust counteracting mechanisms to evade or suppress type I IFN responses [29,30], which could limit the protective efficacy of TRIM29 ablation. Thus, while TRIM29-KO pigs represent a promising model for broad-spectrum antiviral immunity, their defense capabilities are contingent on the specific viral pathogen and its immune evasion strategies.

Although emerging evidence indicates that the down-regulation of TRIM29 plays a protective role in host antiviral immune responses through the activation of innate immunity, an overactive immune system can lead to excessive cytokine production, which has been associated with severe immunological disorders and tissue injury. Notably, studies have reported that Trim29-KO mice exhibit increased susceptibility to LPS-induced inflammation and H. influenzae infection in the lungs, due to the overproduction of pro-inflammatory cytokines [18]. While it is necessary to recognize that TRIM29-KO animals might display heightened vulnerability to specific bacterial infections or environmental stressors, conclusive validation of these hypotheses in pigs demands additional research. Our findings demonstrate that TRIM29-KO animals do not exhibit elevated inflammation under physiological conditions, and that PAMs derived from TRIM29-KO pigs do not express higher levels of pro-inflammatory cytokines upon LPS stimulation. Of note, the TRIM29-KO pigs even displayed reduced basal levels of several pro-inflammatory cytokines, including IL-8, TNFα, IFNγ, and IL17A. The mechanism underlying the decreased basal levels of these pro-inflammatory cytokines warrants further investigation. Combined with the observed growth performance in conventional farming settings, these genetically modified pigs show no overt health abnormalities under normal physiological parameters. However, systematic evaluation of their responses to diverse pathogenic exposures—including bacterial, viral, or inflammatory triggers—is imperative to comprehensively evaluate the biosafety, disease resilience, and agricultural viability of this novel pig breed. Another potential limitation of our modified pigs is that the heightened antiviral response could act as a powerful selective pressure, accelerating viral evolution and promoting the emergence of variants capable of evading this key host defense. This risk underscores the necessity for a thorough and comprehensive investigation of these animals prior to practical application.

Materials and Methods

Ethics statement

The study was performed in compliance with the Guidelines for Ethical and Welfare Review of Laboratory Animals (GB/T 35892–2018), and the Technical Specification for Ethical and Welfare Review of Laboratory Animals (DB11/T 1734–2020) of China. All animal experiments were reviewed and approved by the Institutional Animal Care and Use Committees at South China Agricultural University.

Viruses and cells

The viruses used in this study are a virulent field isolate of PRV (GD-HY, GenBank accession number MT197597), VSV-EGFP engineering strain in Indiana background and TGEV vaccine strain (WH-1R). PK15 cells were from Chinese National Collection of Authenticated Cell Cultures (#GNO31). IPI-FX cells were kindly provided by Professor Yongchang Cao (Sun Yat-sen University).

Viral infection and quantification

PRV and VSV were cultured and titrated in PK15 cells, and TGEV was propagated and quantified in IPI-FX cells. PRV, VSV, and TEGV were inoculated onto confluent cell monolayers at a MOI of 0.1 or 1. The cells were maintained in high-glucose DMEM (Gibco) supplemented with 2% fetal bovine serum (FBS; Gibco) and incubated at 37°C in a 5% CO₂ atmosphere. Upon observation of complete cytopathic effect (CPE), the culture vessels underwent three freeze-thaw cycles to facilitate the release of intracellular viral particles. The viral titers were subsequently determined using the TCID₅₀ method, and the virus solutions were aliquoted and stored as stock preparations at -80°C.

TCID₅₀ or quantitative PCR (qPCR) were performed to quantify the infectious virus titers or virus DNA/RNA levels, respectively. For TCID₅₀ assay, virus was serially diluted 10-fold in DMEM with 2% FBS, and 100 µL of each dilution was inoculated onto PK15 cell monolayers pre-seeded in 96-well plates. Following inoculation, the plates were incubated at 37°C under 5% CO₂ for 1 week. CPE in individual wells were was monitored and recorded. Viral titers, expressed as TCID₅₀/100 μL, were analyzed using the Reed and Muench endpoint calculation method to quantify infectious units. For qPCR analysis of viral DNA or RNA in the tissues and cell culture supernatants, viral nucleic acid was extracted using the RaPure Viral DNA/RNA Kit (Magen Biotech) and quantified using Taq Pro HS Universal Probe Master Mix or HiScript II One Step qRT-PCR Probe Kit (Vazyme) in QuantStudio 7 Flex Real-Time PCR System (Thermo Fisher Scientific). The PRV primers and probe were PRV-gE-Forward, CCCACCGCCACAAAGAACACG, PRV-gE-Reverse, GATGGGCATCGGCGACTACCTG, and PRV-gE-Probe, FAM-CGGCGCGAGCCGCCCATCGTCAC-BHQ1. The VSV primers and probe were VSV-L-Forward, GATACAGTACAATTATTTTGGGAC, VSV-L-Reverse, ATGGCGTATTTGAAAGTAGAA, and VSV-L-Probe, FAM-TTGATGATGCATGATCCTGCTCTTCGT-BHQ1. The TGEV primers and probe were TGEV-S- Forward, TGAGGGTGCTGGCTTTGAT, TGEV-S-Reverse, CAAGAGTGACACCACCCGTT, and TGEV-S-Probe, VIC-CACTGTGGCACCCTTACCTGATTGT-BHQ1. To enable absolute quantification of viral copies, standard curves were generated using plasmids containing the sequences of PRV gE, VSV L, and TGEV S genes. These plasmids served as standards, allowing for the precise calculation of viral copy numbers in the samples based on the standard curves.

TRIM29 knockdown and overexpression

siRNA targeting the coding region of pig TRIM29 mRNA was synthesized and transfected to knock down gene expression. For overexpression, the full-length pig TRIM29 cDNA was amplified by RT-PCR and cloned into the eukaryotic expression vector pcDNA3.1(+). Both the TRIM29 overexpression construct and the empty vector control were purified to achieve endotoxin levels below 10 EU/mg, thereby eliminating potential confounding effects on IFN production. Following 24-h transfection with either siRNA or plasmid, cells were infected with PRV or VSV for a further 24 h. Subsequently, viral titers in the culture supernatant and intracellular gene expression levels were analyzed.

Generation of Trim29-KO mice

Trim29-KO mice in the C57BL/6J genetic background were generated using a double gRNA-mediated CRISPR/Cas9 approach, where two gRNAs targeting the 5'UTR region (GATGGATAGATTGCAGATCCTGG) and the intron following exon 3 (GATAGTGAGTGTATGGTAATGGG) of the Trim29 gene were synthesized and complexed with SpCas9 protein for cytoplasmic microinjection into the pronuclear-stage embryos. After embryo transfer, we obtained F0 generation chimeric mice, which were further bred to produce F1 generation mice containing heterozygous individuals. Genotyping of the F1 mice was performed using two primer pairs, one (CAAACACAGGCAGGTCTGAGCTA and GAGTGGATGGGTGATAGGTGGGC) amplifying the KO allele (406 bp) and the other (CCCTTGTGTTTATTCTCTTATGAG and CTGCTATCACCATCTTGAAGTGTC) amplifying the WT allele (531 bp), confirming a 13,508-bp deletion spanning exon 1 to exon 3 of the Trim29 gene; heterozygous KO mice were then interbred to generate homozygous Trim29-KO mice which completely lack functional Trim29.

Generation of TRIM29-KO pigs

Primary PFFs were isolated from a male Yorkshire pig fetus and transfected with the PX330 plasmid (#42230, Addgene) harboring the spCas9 gene and a gRNA targeting the exon 1 region of pig TRIM29 (GACCTCCAGCTACTTCAGCATGG) using the Nucleofector system (LONZA). Following transfection, the PFFs were cultured at low density in culture dishes to facilitate the formation of single-cell-derived colonies without the application of drug selection pressure. Colonies exhibiting optimal morphology and sufficient cell numbers were carefully selected and expanded in 48-well plates. Upon reaching confluence, a small aliquot of cells from each well was lysed with proteinase K, and the resulting lysates were subjected to PCR amplification to detect the presence of indels at the gRNA-targeted site. Colonies demonstrating biallelic indels at the Cas9 cleavage site were identified and chosen as donor cells for SCNT. SCNT and subsequent embryo transfer were conducted following established protocols as previously described. The pregnancy status of surrogate sows was monitored monthly, and cloned piglets were delivered via natural birth. Genotyping of the TRIM29 gene in the cloned piglets was performed by PCR amplification of the edited region using genomic DNA extracted from ear biopsies as the template.

PRV and VSV challenge in mice

Trim29 homozygous KO mice, heterozygous KO mice, and WT littermates were intraperitoneally injected with 100 μL PRV (1 × 10⁴ TCID₅₀/100 μL). Disease progression and survival rates were monitored for 96 h post-infection. Animals were humanely euthanized once they reached a moribund state. All remaining mice were sacrificed at the experimental endpoint of 96 h post-challenge, and serum, brain and lung tissues were collected for further analysis. Viral DNA was extracted from equal amounts of serum and tissue samples and quantified using the aforementioned method. The mRNA expression levels of IFNα and IFNβ were measured by SYBR Green qPCR, with GAPDH serving as the internal reference gene for normalization.

Mice were infected with VSV via intraperitoneal injection of 1 mL viral suspension (1 × 10⁸ TCID₅₀/100 μL). Following infection, disease progression and survival rates were monitored daily for 12 days. All surviving animals were euthanized on day 12 post-infection, at which point tissue samples and serum were collected for subsequent analysis of viral loads and quantification of IFNα and IFNβ expression levels using the aforementioned methods.

PRV challenge in pigs

TRIM29-KO and WT Yorkshire pigs of the same age and origin were raised in the same group after weaning. All animals received identical diets and remained unvaccinated. To assess the presence of PRV maternal antibodies, PRV gB antibodies were detected monthly from serum samples using both a PRV gB antibody indirect ELISA kit (SK109 PRVgB, Biochek) and a PRV gB antibody competitive ELISA kit (YB034, Wens Group). Additionally, PRV gE antibodies were measured using a PRV gE antibody competitive ELISA kit (YB015, Wens Group) to monitor potential infections by field strains of PRV. Prior to PRV challenge, the pig populations were monitored to confirm the absence of antibodies and antigens of major pig pathogens, including PRRSV, CSFV, African swine fever virus (ASFV), foot-and-mouth disease virus (FMDV), and porcine parvovirus (PPV). At the age of 5 months, when the maternal antibodies against PRV gB had declined to seronegative levels, the pigs were subjected to a PRV challenge. Each animal received an intranasal inoculation of 10 mL PRV culture (1 × 10^6.9 TCID₅₀/100 μL), with 5 mL administered into each nostril. Following inoculation, the clinical symptoms of each pig were monitored and recorded daily. PRV-associated pathological manifestations were assessed using a clinical scoring system: 0 (normal), 1 (hyperthermia with body temperature 40–41°C), 2 (severe fever >41°C combined with respiratory distress), 3 (ataxia), 4 (convulsions), and 5 (terminal moribund state or death) [31,32]. Nasal swabs were collected from all pigs on days 2, 4, 6, and 8 post-infection. PRV DNA extracted from these samples was analyzed via qPCR to monitor viral shedding levels. Blood samples obtained on days 7 and 9 were assayed to determine viremia levels, PRV-specific antibody titers and type I IFN concentrations. Pigs were humanely euthanized at the predefined humane endpoints (moribund state). On day 9, surviving pigs were euthanized and tissue specimens (brain and lung) were harvested for viral load quantification, type I IFN gene expression analysis and histopathological evaluation.

Measurement of viral copies in tissues, serum and swab samples

Equal quantities of tissue and serum specimens were processed for viral DNA extraction. Nasal swabs were vortexed in 2 mL of PBS, and 200 μL of the resulting suspension was subjected to viral nucleic acid extraction using the RaPure Viral DNA/RNA Kit (Magen Biotech) following the manufacturer’s protocol. All extracts were eluted in a final volume of 50 μL per specimen. To establish a standard curve for absolute quantification of viral copies in specimens, a PRV gE plasmid with precisely determined DNA copy numbers was subjected to 10-fold serial dilutions and used as qPCR templates. This dilution series enabled generation of a standard curve by plotting threshold cycle (Ct) values against the corresponding DNA copy numbers. The qPCR analysis was performed as mentioned above and viral load quantification was achieved by extrapolating Ct values against run-specific standard curves.

ELISA

PRV gB and gE antibody levels were quantified using competitive ELISA kits (Wens Group). Serum samples were diluted 1:2 for gB detection and 1:5 for gE detection using the provided assay buffer. Following the manufacturer's protocol, diluted serum samples along with positive and negative controls were loaded into antigen-precoated 96-well microplates. After incubation and washing steps, HRP-conjugated specific antibodies were added to compete with serum antibodies for binding to the immobilized antigens. Following additional washing to remove unbound conjugates, enzymatic reactions were developed using TMB substrate, and absorbance was measured at both 450 nm and 620 nm. The S/N ratio was calculated using the following formula: S/N ratio = [OD450nm (sample) - OD620nm (sample)] / [OD450nm (negative control) - OD620nm (negative control)]. A cutoff value of 0.6 was established for result interpretation, where S/N ratios show an inverse correlation with antibody concentrations in the tested samples.

The concentrations of IFNɑ and IFNβ in pig serum were measured using a pig IFNɑ ELISA Kit (47100, PBL) and a porcine IFNβ ELISA Kit (ES8RB, Thermo Fisher Scientific), respectively, following a similar protocol. Briefly, Serum samples were diluted (1:10 for IFNα, 1:2 for IFNβ) and added to 96-well microplates pre-coated with capture antibodies. Serially diluted standards were included on the same plate to generate standard curves. After incubation and washing, biotin-conjugated detection antibodies were added, followed by reaction with streptavidin-HRP and color development using TMB substrate. The reaction was then stopped, and the absorbance at 450 nm was measured. Standard curves were constructed using the four-parameter logistic (4PL) model, and the concentrations of IFNɑ and IFNβ in the serum samples were calculated based on the standard curves.

Pig IgG, IgA and IgM levels were quantified using a competitive inhibition enzyme immunoassay with pig-specific ELISA kits: IgG (CSB-E06804p, CUSABIO), IgA (CSB-E13234p), and IgM (CSB-E06805p). Following differential dilutions (IgG: 1:1000, IgA/IgM: 1:2000), diluted serum samples and standards were loaded onto antigen-precoated microplates. Subsequently, HRP-conjugated secondary antibodies were added, and plates were incubated at 37°C for 40 min. After incubation, plates underwent five wash cycles, followed by TMB substrate addition for chromogenic reaction. Optical density was measured at 450 nm using a microplate reader, and analyte concentrations were determined via 4PL regression based on standard curves.

Quantitative PCR

To assess the gene expression levels of IFNɑ and IFNβ in animal tissues, total RNA was extracted from tissue samples using TRIzol reagent (Thermo Fisher Scientific), followed by cDNA synthesis using HiScript III All-in-one RT SuperMix Perfect for qPCR (Vazyme). The mRNA expression levels of IFNɑ and IFNβ were quantified by qPCR using Taq Pro Universal SYBR qPCR Master Mix (Vazyme) on the QuantStudio 7 Flex Real-Time PCR System, usng GAPDH employed as an internal reference gene for normalization. All reactions were performed in triplicate to ensure reproducibility. Relative gene expression levels were determined using the 2 − ΔΔCT method.

Western blot

Cells were lysed using Pierce RIPA buffer (Thermo Fisher Scientific) supplemented with the Halt protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific). Following centrifugation to clarify the cell lysates, protein concentrations were quantified using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein lysates were mixed with loading buffer, boiled, and separated by SDS-PAGE. The resolved proteins were then transferred onto a PVDF membrane, which was subsequently blocked with 5% non-fat milk and incubated with the corresponding primary antibodies at 4 °C overnight. After washing, the membrane was incubated with an HRP-conjugated secondary antibody. Following thorough washing, the target proteins were visualized using the SuperSignal West Pico Enhanced Chemiluminescence Kit (Thermo Fisher Scientific). The following primary antibodies used in this study, anti-TRIM29 (1:1000; 17542–1-AP; Proteintech), anti-cGAS (1:2000; 26416–1-AP; Proteintech), anti-MAVS (1:2000; 14341–1-AP; Proteintech), anti-STING (1:1000; 19851–1-AP; Proteintech), anti-TBK1 (1:1000; #3504; Cell Signaling Technology), anti-phospho TBK1 (1:1000; #5483; Cell Signaling Technology), anti-IRF3 (1:5000; 11312–1-AP; Proteintech), anti-phospho IRF3 (1:1000; 29528–1-AP; Proteintech), anti-IL-1β (1:500; P420B; Thermo Fisher Scientific), anti-COX2 (1:1000; EPR12012; Abcam), anti-TNFα (1:500; GTX110520; GeneTex), anti-β-actin (1:5000; 4970; Cell Signaling Technology) and anti-GAPDH (1:5000; 2118; Cell Signaling Technology).

Histological Analysis

The tissue samples were fixed in 4% paraformaldehyde for 3 days and subsequently transferred into PBS containing NaN₃ for long-term preservation. For histological analysis, the fixed tissues were dehydrated through a graded ethanol series, embedded in paraffin wax, and sectioned into thin slices. For hematoxylin and eosin (H&E) staining, the tissue sections were deparaffinized in xylene and rehydrated through a descending ethanol series. Subsequently, the sections were stained with hematoxylin and eosin Y to visualize nuclear and cytoplasmic structures, respectively. For immunohistological analysis, the rehydrated tissue sections underwent antigen retrieval in 0.01 M citrate buffer and were then incubated with PRV polyclonal antibody (PA1–081, Thermo Fisher Scientific). The immunostaining signals were developed using an HRP-conjugated Rabbit IgG SuperVision Assay Kit (Boster) with 3,3’-diaminobenzidine (DAB) as the chromogen.

Luminex multiplex immunoassay of cytokines

The selected mouse and pig cytokines and mouse antibody isotypes were determined in serum using the Luminex 200 Instrument System (Thermo fisher Scientific). We used ProcartaPlex Porcine Cytokine & Chemokine Panel 1, 9plex (EPX090-60829-901, Thermo fisher Scientific) to measure pig IFNɑ, IFNγ, IL-1β, IL-10, IL-12, IL-4, IL-6, IL-8 and TNFα levels; Bio-Plex Pro Mouse Cytokine Th17 Panel A 6-Plex assay (M6000007NY, Bio-Rad) to measure mouse IFNγ, IL-1β, IL-6, IL-10, IL-17 and TNFα levels; and ProcartaPlex Mouse Antibody Isotyping Panel 1, 7plex (EPX070-20815-901, Thermo fisher Scientific) to measure IgA, IgE, IgG1, IgG2a, IgG2b, IgG3 and IgM levels. For multiplex assays, capture beads were first added to each well, followed by the addition of standards or samples. The plates were then sealed and incubated overnight at 4°C. After incubation, the beads were sequentially incubated with the biotinylated detection antibody mix followed by streptavidin-PE, with wash steps performed between each incubation. Finally, the beads were resuspended in assay buffer and analyzed on the Luminex reader.

PAMs isolation

PAMs were isolated using lung lavage. Briefly, pigs were euthanized and the trachea was aseptically exposed and cannulated. Sterile, ice-cold PBS was instilled into the lung lobes in aliquots. The lavage fluid was gently massaged within the lungs and subsequently recovered by gravity drainage into sterile collection tubes on ice. The pooled lavage fluid was centrifuged at 500 × g for 10 min at 4°C. The resulting cell pellet was washed twice with PBS, resuspended in RPMI 1640 complete medium, and counted using a hemocytometer with trypan blue exclusion to assess viability. Cells were then aliquoted and cryopreserved in a suitable freezing medium for long-term storage.

LPS-induced inflammation

LPS from Escherichia coli O111:B4 (L2630, Sigma) was used to induce inflammation in PAMs. PAMs seeded in 24-well plates were treated with 20 ng/mL LPS for 12 h prior to harvest. Total RNA was extracted to analyze the mRNA expression levels of TNF-α, IL-6, IL-1β, and COX2. In parallel, PAMs were lysed for protein extraction to assess the protein expression of pro-TNF-α, pro-IL-1β, and COX2.

Off-target analysis

We utilized two computational tools, Cas-Offinder and CRISPOR, to predict potential off-target sites [33,34]. For Cas-Offinder, the parameters were configured as follows: the PAM type was set to SpCas9 from Streptococcus pyogenes (5’-NGG-3’), up to three mismatches and one bulge were allowed, and the total number of mismatches plus bulges was restricted to four. For CRISPOR, the PAM type defined as 20 bp-NGG for SpCas9, SpCas9-HF1, and eSpCas9 1.1, and up to four mismatches permitted. By integrating the predictions from both tools, we compiled a comprehensive list of potential off-target sites for subsequent experimental validation.

To empirically evaluate off-target effects at the whole-genome level, we conducted WGS of the TRIM29-KO pig using PE150 mode in BGISEQ platform. Raw sequencing data underwent quality control with fastp, applying filtering thresholds of a base quality ≥ 30, an average read quality ≥ 30, a maximum of five N bases per read, and removal of polyG tails. High-quality reads were aligned to the reference genome (GCF_000003025.6_Sscrofa11.1.fa) using BWA, and the resulting alignments were processed and converted using SAMtools. PCR duplicates were marked using the MarkDuplicates module in Picard. Variants, including SNPs and InDels, were called using GATK's HaplotypeCaller and filtered based on stringent criteria: QUAL ≥ 60.0, QD ≥ 20.0, FS ≤ 13.0, MQ ≥ 30.0, MQRankSum ≥ -1.65, and ReadPosRankSum ≥ -1.65. The filtered variants were normalized using bcftools and annotated with VEP. Finally, we intersected the predicted off-target sites from Cas-Offinder and CRISPOR with the variants identified through WGS analysis to generate a list of candidate off-target sites. These candidates were individually PCR-amplified and sequenced to validate mutations against the wild-type genome.

Statistical Analysis

All data were expressed as mean ± s.d. and plotted using GraphPad Prism. The Kolmogorov–Smirnov test was performed to determine whether the data were normally distributed. For analysis of normally (parametric) distributed data, the mean values from individual groups were compared using either an independent sample unpaired Student’s t-test (for equal variances), Welch’s t-test (for unequal variances), one-way analysis of variance (ANOVA) or repeated measurements of two-way ANOVA with a post hoc test and P values adjusted for multiple comparisons, using statistical hypothesis testing or by controlling the False Discovery Rate, dependent on sample sizes (For 2 < n < 6, using Sidak's multiple comparisons test; for n > 6, using two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli for multiple comparisons). For non-normally (nonparametric) distributed data, the median values from individual groups were compared with Mann Whitney test (comparisons of 2 groups) and Kruskal-Wallis test followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli for multiple comparisons (comparisons of 3 or more groups). Survival curve between groups were analyzed using the Gehan-Breslow-Wilcoxon test. The exact p values, adjusted p values, or FDR-corrected q values depending on different statistical analysis are described in each figure and between the indicated groups. The level of significance was set at 0.05, and trends were discussed at 0.1.

Supporting information

S1 Table. Predicted off-target sites in the TRIM29-KO pigs.

(XLSX)

S1 Fig. Graphical abstract.

(PNG)

Data Availability

The WGS data of TRIM29-KO pigs have been deposited in the NCBI Sequence Reads Archive (SRA) database under BioProject PRJNA1233636. All relevant data generated in this study are present within the manuscript and supplemental information.

Funding Statement

This research was financially supported through the following funding sources: the Guangdong Provincial Agricultural Breeding Industry Development Program administered by the Department of Agriculture and Rural Affairs of Guangdong Province (2024-XPY-00-015 to HY), the National Science and Technology Major Project for Biological Breeding (2023ZD0404303 to HY), and the General Program of the National Natural Science Foundation of China (32372874 to HY). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.