Nanobody-based bioPROTAC for viral protein degradation provides an antiviral strategy for porcine arterivirus

Abstract

Background

Proteolysis-targeting chimeras (PROTACs) are powerful tools for targeted protein degradation and are expected to contribute to a promising strategy for next-generation precision therapeutic antiviral drug development. Nanobody-based bioPROTACs can directly bind to protein and mediate target protein degradation, providing a potential antiviral strategy for RNA viruses featuring error-prone replication. Here, we aimed to establish a modular speckle-type POZ protein (SPOP)-derived bioPROTAC platform that enabled rapid antiviral drug construction through the substitution of a target protein-specific nanobody.

Results

Using porcine reproductive and respiratory syndrome virus (PRRSV) as a model pathogen, bioPROTACs molecules were successfully constructed by flexibly fusing nanobodies against PRRSV nonstructural protein 9 (Nsp9, viral RdRp) to the BTB domain of SPOP. BioPROTACs demonstrated specific degradation of target proteins in a dose-dependent manner, and a bivalent nanobody configuration enhanced the degradation efficiency to greater than 60%. BioPROTACs exhibited antiviral activity against multi-lineages of PRRSV and significantly potentiated the antiviral efficacy of non-neutralizing nanobodies in vitro. Furthermore, intravenous delivery of bioPROTAC-encoding constructs in mice achieved significant reduction of target protein levels within 24 h, demonstrating efficient in vivo degradation capability. Moreover, the combined administration of bioPROTACs via the mRNA-LNP system suppressed PRRSV proliferation and transmission in piglets, which was characterized by reduced viremia, alleviated lung damage, and a decrease in the piglet mortality rate to 25%. Importantly, we revealed that the subcellular localization of both the target protein and bioPROTACs determined the degradation pathway, confirming that cytoplasmic 9nb-SPOP_ΔNLS mediated Nsp9 degradation through the autophagy–lysosome pathway in the cytoplasm. This study expands the applicability of SPOP-derived bioPROTACs from nuclear proteins to cytoplasmic proteins, providing a novel strategy for developing antiviral therapies against highly variable viruses.

Conclusion

The aim of the current study was to develop and validate modular bioPROTACs targeting essential viral proteins. We constructed the degraders by fusing target-specific nanobodies to the BTB domain of SPOP. More importantly, a combination of bioPROTACs targeting different stages of viral replication, delivered via mRNA-LNPs, suppressed viral replication in a pig model. These findings offer valuable insights into the target degradation mechanisms of SPOP-derived bioPROTACs and provide a foundation for the design of antivirals that have activity against multi-lineages of porcine arterivirus and overcome drug resistance.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04348-8.

Introduction

Viral infectious diseases pose a formidable global health challenge and are characterized by rapid transmission, broad impact, and significant mortality. Currently, the combined use of vaccination and antiviral medication serves as the primary strategy for prevention and control [1]. However, the effectiveness of this approach is often limited. Vaccination occasionally fails to confer robust protection against highly variable RNA viruses, while the use of existing antiviral drugs is increasingly compromised by the emergence of drug-resistant strains [1–3]. Therefore, the development of novel antiviral agents with innovative mechanisms of action, which are crucial for reducing drug resistance and achieving broader-spectrum activity, is urgent.

Targeted protein degradation (TPD) is a promising therapeutic approach with the potential to treat noninfectious or infectious diseases via protein degradation [4–6]. Recently, many different TPD tools have been developed, including proteolysis-targeting chimeras (PROTACs), molecular glues, antibody-based PROTACs (AbTACs) and lysosome-targeting chimeras (LYTACs) [7, 8]. PROTAC technology represents a therapeutic strategy in which target proteins are selectively degraded by the intracellular protein degradation machinery. A typical PROTAC molecule consists of three key components: a ligand that binds to the protein of interest, a ligand that recruits an E3 ubiquitin ligase, and a linker connecting the two ligands. As a heterobifunctional molecule, PROTAC facilitates the ubiquitination and subsequent degradation of the target protein by bringing the E3 ubiquitin ligase into proximity with the target [9]. The specific E3 ubiquitin ligase recruited by a PROTAC molecule directly determines the catalytic machinery engaged for target ubiquitination, thereby critically influencing both the degradation efficiency and selectivity of the degrader.

Although more than 600 E3 ligases have been identified in humans, only limited E3 ligands, such as the VHL and CRBN ligands, are frequently used to design PROTAC degraders [10]. Moreover, the variety of ligands available for biosynthesis is also limited. The recent discovery and validation of several novel E3 ligase ligands have significantly broadened the diversity of available E3 recruiters. Speckle-type POZ protein (SPOP) acts as a novel cullin 3-RING ubiquitin ligase (CRL3) adaptor, enabling the degradation of target endogenous nuclear proteins via the PROTAC strategy [11, 12]. SPOP functions as a substrate receptor within the CRL3 complex, recognizing specific proteins and delivering them for ubiquitination and proteasomal degradation. SPOP comprises three distinct domains: an N-terminal meprin and TRAF-C homology (MATH) domain responsible for substrate recognition; a central broad-complex, tramtrack and bric-a-brac (BTB) domain that mediates SPOP dimerization and recruits cullin 3 (Cul3), which is the scaffold protein of CRL3 ligase; and a C-terminal BTB and C-terminal Kelch (BACK) domain that provides a secondary dimerization interface [13]. Notably, SPOP oligomers can undergo liquid–liquid phase separation (LLPS) with their substrates to form dynamic, droplet-like membrane-less organelles, also called biomolecular condensates [14]. This condensate formation facilitates the efficient enrichment of the CRL3 ligase into these compartments, thereby enhancing the efficiency of substrate ubiquitination [15]. These features make SPOP an attractive E3 ligase ligand in PROTAC applications.

To date, various PROTAC molecules have been developed as degradation drivers for viral targets by utilizing known chemical ligands, with successful examples including SARS-CoV-2 M-pro (3CLpro), HCV NS3/4A protease, influenza A virus neuraminidase and HIV-1 Vif [16–19]. However, for many viruses that currently lack effective small-molecule inhibitors, the discovery of effective heterobifunctional compounds remains challenging, significantly impeding the development of new antivirals against these infectious diseases.

Recently, nanobody-conjugated PROTACs (bioPROTACs) have been developed, and these chimeric molecules have demonstrated efficient degradation of the target protein [11, 20, 21]. A nanobody is a single-chain antibody fragment, also known as a VHH antibody, derived from the variable-structure domain of a camel heavy-chain antibody (HcAb). Conventional antibody drugs rely on active-site blockade or viral neutralization, while nanobody-based bioPROTACs degrade the protein of interest (POI) through binding to noncatalytic sites, thereby reducing the risk of drug resistance. Compared with the conventional small-molecule ligands that are commonly used in PROTACs, nanobodies exhibit remarkable specificity and affinity for protein antigenic sites. In addition, multivalent or multispecific nanobodies can be prepared to optimize binding affinity by modifying the encoding genes in complementarity-determining regions (CDRs) [22, 23]. It has been reported that the fusion of a nanobody to SPOP generates a bioPROTAC that depletes nuclear BCL11A, which is a critical repressor of fetal globin gene transcription and has historically been intractable to conventional small-molecule inhibitors [24]. Although SPOP is a nuclear E3 ubiquitin ligase, hypoxic conditions induce its cytoplasmic accumulation in clear cell renal cell carcinoma. Cytoplasmic SPOP promotes the ubiquitination and degradation of the tumor suppressor PTEN, thereby driving tumorigenesis [25]. How SPOP functions in the cytoplasm and whether it can be engineered into a general platform for degrading cytoplasmic proteins, such as cytoplasmic localized viral proteins, remain unknown. However, for viruses that lack known small-molecule inhibitors, nanobodies and SPOP provide ideal ligands for constructing PROTACs with high specificity and affinity.

Porcine reproductive and respiratory syndrome virus (PRRSV) is an enveloped, single-stranded positive RNA virus that is widespread on swine farms worldwide and causes enormous economic losses [26]. Owing to the high mutation and recombination rates between circulating strains, PRRSV is an ideal model for evaluating antivirals against mutation-prone viruses. Nonstructural protein 9 (Nsp9), which possesses essential RNA-dependent RNA polymerase (RdRp) activity for viral RNA synthesis, and the nucleocapsid (N) protein play critical roles in different stages of the viral life cycle [27, 28]. Studies have shown that nanobodies targeting the PRRSV Nsp9, Nsp4 and N proteins significantly suppress viral replication in vitro. Fc-fused nanobodies against PRRSV Nsp9 and N protein enhance delivery to porcine alveolar macrophages and improve antiviral efficacy in vivo [29–31]. Furthermore, a multitarget nanobody-peptide conjugate targeting PRRSV N, Nsp9 and cellular CD163 was shown to exert synergistic anti-PRRSV effects by blocking multiple stages of viral replication [32]. Thus, targeting key PRRSV proteins and different replication stages are promising therapeutic strategies. In addition, they are highly conserved in North American-type (PRRSV-2) and European-type (PRRSV-1) strains, suggesting their potential as promising multi-lineages of PRRSV antiviral targets. However, conventional small-molecule inhibitors targeting enzymatic activities or protein–protein interfaces are often susceptible to rapid resistance because of the high mutation rate of RNA viruses [33]. This phenomenon highlights the critical need for antivirals that are less prone to generating drug-resistant strains. Therefore, we hypothesize that by fusing nanobodies targeting Nsp9/N with E3 ligase ligands, these key viral proteins could be efficiently degraded, thereby providing an antiviral strategy and decreasing the susceptibility to drug resistance.

BioPROTACs are characterized by high molecular weight (> 700 Da), substantial polar surface area, and multiple hydrogen bond donors, which lead to inadequate cell permeability, poor solubility, and restricted oral bioavailability [34, 35]. These delivery limitations not only reduce degradation efficiency but also raise concerns over potential off-target effects at high doses [36]. Currently, nanomedicine delivery systems have emerged as pivotal platforms in biomedical research and clinical therapy owing to their remarkable drug-loading capacity, efficient delivery performance, and controllable release profiles [37–39]. Among them, liposomal nanoparticles (LNPs) represent a well-established system featuring excellent biocompatibility, biodegradability, low immunogenicity, and the ability to protect encapsulated drugs from degradation. Furthermore, LNPs can deliver multiple therapeutic agents to synergistically increase therapeutic efficacy [40, 41]. These characteristics establish LNPs as a highly promising delivery platform for PROTAC-based therapeutics.

Given the limitations of current antivirals against rapidly mutating viruses such as PRRSV, the aim of this study was to develop and validate modular bioPROTACs targeting essential viral proteins. We constructed these degraders by fusing target-specific nanobodies to the BTB domain of SPOP. More importantly, a combination of bioPROTACs targeting different stages of viral replication, delivered via mRNA-LNP, suppressed viral replication in a pig model. These findings offer valuable insights into the target degradation mechanisms of SPOP-derived bioPROTACs and provide a foundation for the design of antivirals that suppress viral replication and overcome drug resistance.

Materials and methods

Cells and viruses

Human embryonic kidney (HEK293T) cells and monkey kidney (Marc-145) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA). The PRRSV strains used in this study were isolated and preserved in our laboratory and included three PRRSV-2 strains, the HP-PRRSV HuN4 strain (GenBank accession number: EF635006.1), the NADC30-like PRRSV HLJWK108-1711 strain (GenBank accession number: MN046230.1) [42], and the NADC34-like PRRSV LNTZJ1341-2012 strain (GenBank accession number: OL516360.1 [43], and one PRRSV-1 strain, ZD1 [44]. All the PRRSV strains were propagated and titrated in Marc-145 cells maintained in DMEM supplemented with 2% FBS. EGFP-labeled PRRSV (HuN4-EGFP) was kindly provided by Prof. Fei Gao, Shanghai Veterinary Research Institute, Chinese Academy of Agricultural Sciences, as described in a previous study [45].

Construction of plasmids

All plasmids expressing PRRSV Nsp9, N protein, and bioPROTAC molecules were constructed as described below. The PRRSV Nsp9 and N genes cloned from the HP-PRRSV strain HuN4, verified by sequencing, were finally subcloned and inserted into the EcoR I/Xho I sites of the pCAGGS vector by T4 DNA Ligase (Takara, Japan) to generate the recombinant pCAGGS-HA-Nsp9 and pCAGGS-HA-N plasmids. The nanobodies against PRRSV Nsp9 (Nsp9nb, GenBank accession number: PP779760) and PRRSV N (Nb1, named as Nnb) were preserved in our laboratory [46], and the fragments of Nsp9nb and Nnb were amplified from the pCAGGS-HA-Nsp9nb and pCAGGS-HA-Nnb plasmids with 5’ BamH I and 3’ Kpn I restriction recognition sites. The sequence of irrelevant nanobody (NCnb) was provided by our laboratory. This nanobody did not react with the PRRSV Nsp9 or N protein and did not demonstrate an inhibitory effect on the replication of PRRSV; therefore, NCnb was used as a negative control in this study. The cDNA sequence encoding porcine SPOP167-374 (GenBank accession number: XM_021067137.1) was amplified by PCR from the genome of porcine alveolar macrophages with 5’ Kpn I and 3’ Xho I restriction recognition sites. The fragments mentioned above were subsequently subcloned and inserted into the pcDNA3.1(+), pLEX-IRES-mCherry, and pGEX-6P-1 plasmids with BamH I and Xho I, respectively, using a ClonExpress Ultra One Step Cloning Kit (Vazyme, China). All the constructs were verified by gene sequencing. The primers used for plasmid construction are listed in Table 1.

Table 1.

Primers used for plasmid construction

Primer/probe name	Sequence (5’-3’)	Purpose
Nsp9-F	GTTCCAGATTACGCTGAATTCCTAGCCGCCAGCGGCT	pCAGGS-HA-Nsp9
Nsp9-R	AAGATCTGCTAGCTCGAGTTACTCATGATTGGACCTG
Nsp9 L6F F	AGCGGCTTGACCCGCTGTGGTCGCGGCGGC	pCAGGS-HA-Nsp9FT
Nsp9 L6F R	ACAGCGGGTCAAGCCGCTGGCGGCTAG
Nsp9 A612T F	TTGGGATAGCGCAGTGCGCCCGCAAGGAC
Nsp9 A612T R	GCACTGCGCTATCCCAACCACGAGCTCTT
N-F	GTTCCAGATTACGCTGAATTCCCAAATAACAACGGCAAGCA	pCAGGS-HA-N
N-R	AAGATCTGCTAGCTCGAGCTGAGGGTGATGCTG
Nsp9-mCherry-F	ATCATGAGGGATCCGTGAGCAAGGGCGAGGAGGAT	pCAGGS-HA-Nsp9-mCherry
Nsp9-mCherry-R	TGCTAGCTCGAGCTTGTACAGCTCGTCCATGC
N-mCherry-F	ACCCTCAGCAGGATCCGTGAGCAAGGGCGA	pCAGGS-HA-N-mCherry
N-mCherry-R	TAGCTCGAGGCTTGTACAGCTCGTCCATGC
3.1-F	TAAGCTTGGTACCGAGCTCGGATCCATGGATTACAAGGATGACGA	pcDNA3.1-FLAG-9nb-SPOP
3.1-SPOP-R	AAACGGGCCCTCTAGACTCGAGTTAGGATTGCTTCAGGCGTT
3.1-9nb-R	AAACGGGCCCTCTAGACTCGAGTTATGAGGAGACGGTGACCTGGG	pcDNA3.1-FLAG-9nb
3.1-NCnb-SPOP-F	TTGGTACCGAGCTCGGATCCATGGATTACAAGGATGACGACGATAAGCTGCAGATGGCCGAGGTGCA	pcDNA3.1-FLAG-NCnb-SPOP
3.1-△NLS-R	AAACGGGCCCTCTAGACTCGAGTTATGCTGAAGCCAGAGAA	pcDNA3.1-FLAG-9nb-SPOP△NLS
3.1-Nnb-SPOP-F	GGATCCATGGATTACAAGGATGACGACGATAAGGCCACCA	pcDNA3.1-FLAG-Nnb-SPOP
3.1-Nnb-SPOP-R	TGTTGACGGAACTTCCACCTCCACCACTTCCACCTCCACCTGAGGAGACGGTGACCTGGG
3.1-mut-F	TGGTGGAGGTGGAAGTTCCGTCAACATTTCTGGCCA	pcDNA3.1-FLAG-9nb-SPOPmut
3.1-mut-R	TGGTGGAGGTGGAAGTTCCGTCAACATTTCTGGCCA
3.1-mut△NLS-R	AAACGGGCCCTCTAGACTCGAGTTATGCTGAAGCCAGAGAA	pcDNA3.1-FLAG-9nb-SPOPmut△NLS
pLEX-F	ACCGACTCTACTAGAGGATCCATGGATTACAAGGATGACGACGA	pLEX-FLAG-9nb-SPOP
pLEX-R	TTCGGCCAGTAACGTTAGGGGGGGGGGAGGGAGAGGGGCGGATCCTTAGGATTGCTTCAGGCGTTTG
pLEX-△NLS-R	TTCGGCCAGTAACGTTAGGGGGGGGGGAGGGAGAGGGGCGGATCCTTATGCTGAAGCCAGAGAACGGTA	pLEX-FLAG-9nb-SPOP△NLS
9nb-F	TTCCAGGGGCCCCTGGGATCCATGGATTACAAGGATGACGACGATAAGGCCACCATGCAGGTCCAACT	pGEX-6P-1-FLAG-9nb
9nb-R	ACGATGCGGCCGCTCGAGGGTACCTGAGGAGACGGTGACCTGGG
SPOP-F	ACCGTCTCCTCAGGTACCTCCGTCAACATTTCTGGCCA	pGEX-6P-1-FLAG-pre-9nb-SPOP/ pGEX-6P-1-FLAG-9nb-SPOP/ pGEX-6P-1-FLAG-NCnb-SPOP
SPOP-R	ACGATGCGGCCGCTCGAGTTAGGATTGCTTCAGGCGTT
RNF4-RING-F	AGGTCACCGTCTCCTCAGGTACC GGTGGAGGTGGAAGTGGTGGAGGTGGAAGTAGTTGTCCGATCTGCATGGA	pGEX-6P-1-FLAG-RNF4-RING
RNF4-RING-R	AGTCAGTCACGATGCGGCCGCTCGAGTCATATATAAATGGGGTGGTA
3.1-F	TAAGCTTGGTACCGAGCTCGGATCCATGGATTACAAGGATGACGA	pcDNA3.1-FLAG-RNF4-RING
3.1-RING-R	TAAACGGGCCCTCTAGACTCGAGTCATATATAAATGGGGTGGT
VN-F	AGTTAAGCTTCTCGAGGCCACCATGTACCCATACGATGT	pVN-Nsp9
VN-R	CTTGCTCACTCCGGATCCACCTCCACCCTCATGATTGGACCTGAGTT
VC-9nb-F	AGTTAAGCTTCTCGAGATGGATTACAAGGATGACGA	pVC-9nb
VC-9nb-R	TCTTCTGCTTGTCGGCACTTCCACCTCCACCTGAGGAGACGGTGACCTGGGT
VC-SPOP-R	TCTTCTGCTTGTCGGCACTTCCACCTCCACCGGATTGCTTCAGGCGTTTGCGT	pVC-9nb-SPOP
VC-SPOP△NLS-R	TCTTCTGCTTGTCGGCACTTCCACCTCCACCTGCTGAAGCCAGAGAACGGTA	pVC-9nb-SPOP△NLS
VC-SPOPmut△NLS-R	TCTTCTGCTTGTCGGCACTTCCACCTCCACCTGCTGAAGCCAGAGAACGGTA	pVC-9nb-SPOPmut△NLS
VC-NCnb-R	TCTTCTGCTTGTCGGCACTTCCACCTCCACCGGATTGCTTC	pVC-NCnb-SPOP
28a-Nsp9-F	AGCAAATGGGTCGCGGATCCCTAGCCGCCAGCGGCTTGAC	pET-28a-His-Nsp9
28a-Nsp9-R	TGGTGGTGGTGGTGCTCGAGCTCATGATTGGACCTGAGTTT
28a-Nsp9-mCherry-F	ATCATGAGGGATCCGTGAGCAAGGGCGAGGAGGAT	pET-28a-His-Nsp9-mCherry
28a-Nsp9-mCherry-R	TGCTAGCTCGAGCTTGTACAGCTCGTCCATGC
28a-N-mCherry-F	ACCCTCAGCAGGATCCGTGAGCAAGGGCGA	pET-28a-His-N-mCherry
28a-N-mCherry-R	TAGCTCGAGGCTTGTACAGCTCGTCCATGC
28a-9nb-SPOP△NLS-F	TTGGTACCGAGCTCGGATCCATGGATTACAAGGATGACGA	pET-28a-9nb-SPOP△NLS
28a-9nb-SPOP△NLS-R	TGGTGGTGGTGCTCGAGTGCTGAAGCCAGAGAACGG
28a-ΔNLS-Fc-F	TGGCTTCAGCACCTATCTGCCCAGGCTGTG	pET-28a-9nb-SPOP△NLS-Fc
28a-ΔNLS-Fc-R	CTCGAGTTACTTCCCCTGAGTCTTAGAGA
28a-9nb-Fc-F	GATTACAAGGATGACGACGATAAGGCCA	pET-28a-9nb-Fc
28a-9nb-Fc-R	GCAGATAGGTGAGGAGACGGTGACCTGGGT
28a-Nnb-F	TATACCATGGATTACAAGGATGACGACGA	pET-28a-Nnb-Fc
28a-Nnb-R	GCAGATAGGTGAGGAGACGGTGACCTGGGT
28a-Nnb-SPOP-R	GCAGATAGGGGATCCGGATTGCTTCAGG	pET-28a-Nnb-SPOP-Fc
28a-NCnb-SPOP-R	GCAGATAGGGGATCCGGATTGCTTCAGG	pET-28a-NCnb-SPOP-Fc
LTV-F	GAACCCGCCACCATGGATGATTACAAGGATGACGACGA	LTV-Nnb-SPOP
LTV-Nnb-SPOP-R	GAGGCTCCAGCTCATTAGATGGATTGCTTCAGGCGTT
LTV-9nb-SPOP△NLS-R	GAGGCTCCAGCTCATTAGATTGCTGAAGCCAGAGAACGGT	LTV-9nb-SPOP△NLS
PRRSV-1 ORF6F	CTGTGAGAAAGCCCGGACT	RT-qPCR
PRRSV-1 ORF6R	GGCCATACTTGACGAGGTTA
PRRSV-1 ORF6 probe	FAM-TGGGCGGCAAYCGAGCTGT-MGB
PRRSV-2 ORF6F	TTGTGCTTGCTAGGCCGC
PRRSV-2 ORF6R	ACGACAAATGCGTGGTTATCA
PRRSV-2 ORF6 probe	FAM-TCTGGCCCCTGCCCA‐MGB

Degradation efficiency by Western blot analysis

HEK293T cells were seeded at a density of 2 × 10⁵ cells per well in 6-well plates, and the cells were cotransfected with pCAGGS-HA-Targets (Nsp9/N) and pcDNA3.1-Flag-bioPROTACs (9nb-SPOP/Nnb-SPOP) at different ratios using X-tremeGENE HP DNA Transfection Reagent (Roche, Germany) according to the manufacturer’s instructions when the cells reached 90% confluency. At 24 h posttransfection (hpt), the cell lysates were collected for western blotting. Briefly, total protein was extracted from cells and quantified with an Enhanced BCA Protein Assay Kit (Beyotime, China). Afterward, the samples were heated at 100 °C for 5 min with loading buffer and separated by 12% SDS–PAGE. After the proteins were transferred from the gel to a PVDF membrane (Millipore, USA), the membrane was blocked with 5% skim milk for 1 h at room temperature and then incubated with the indicated primary antibodies overnight at 4 °C. After three washes with PBST (0.05% Tween-20 in PBS), the membranes were incubated with the secondary antibody for 1 h at room temperature. The signal was captured using a near-infrared fluorescence scanning imaging system. Visualization analysis was performed with Image Studio Lite software, and protein abundance was quantified using ImageJ software. The primary antibodies used were mouse anti-HA tag mAb (1:10,000 dilution; Sigma–Aldrich, USA), rabbit anti-HA tag mAb (1:10,000 dilution; Proteintech, USA), mouse anti-Flag tag mAb (1:5,000 dilution; Sigma–Aldrich, USA), mouse anti-6×His tag mAb (1:5,000 dilution; Proteintech, USA), and mouse anti-β-actin mAb (1:10,000 dilution; Sigma–Aldrich, USA). The secondary antibodies, which were diluted 1:10,000 in PBS, were DyLight 800-labeled goat anti-mouse IgG (KPL, USA) or DyLight 800-labeled goat anti-rabbit IgG (KPL, USA).

Nsp9 and 9nb-SPOP/9nb-RNF4 engagement assay

For the GST pulldown assay, the GST-Flag-9nb-RNF4-RING fusion protein and GST protein were expressed in E. coli BL21(DE3), and the GST tag proteins were purified using Glutathione Sepharose 4B resin. HEK293T cells were transfected with pCAGGS-HA-Nsp9, and the cell lysates were collected at 24 hpt. Afterward, 50 µg of each purified GST-Flag-9nb-RNF4-RING and GST protein was mixed with 200 µL of cell lysates of HA-Nsp9 and 50 µL of glutathione–agarose beads for 4 h at 4 °C with end-over-end mixing. After centrifugation at 2000 × g for 5 min, the supernatants were removed, and the beads were washed 3 times with PBS. Finally, the bead-bound proteins were added to 5× loading buffer (NCM Biotech, China), boiled at 100 °C for 10 min, and analyzed by Western blot analysis.

For the coimmunoprecipitation (Co-IP) assay, HEK293T cells grown to approximately 70% to 80% confluence in 6-well plates were transfected with 1 µg of pCAGGS-HA-Nsp9 and 3 µg of pcDNA3.1-Flag-9nb-SPOP/9nb-RNF4-RING/9nb-RNF4-(RING)₂ per well. At 24 hpt, the cells were lysed with 200 µL of lysis buffer (RIPA with 1% PMSF; Beyotime Biotechnology, China) for 1 h on ice. The cell lysates were centrifuged at 12,000 × g for 10 min, and 50 µL of the supernatant was collected as the input sample. The remaining supernatant was incubated with 20 µL of FLAG M2 beads (Sigma, USA) overnight at 4 ℃. The beads were then centrifuged at 2000 × g for 5 min at 4 ℃ and washed three times with PBS (pH = 7.4). The precipitates were eluted in 40 µL of lysis buffer and 5× loading buffer by boiling at 100 °C for 10 min and analyzed by Western blot analysis.

Activation of Nsp9 ubiquitination

In the ubiquitination assay, HEK293T cells at 90% confluency were cotransfected with the plasmids FLAG-ub, HA-Nsp9, and Flag-Nsp9nb-SPOP using X-tremeGENE HP DNA Transfection Reagent. After 24 hpt, the cells were lysed with lysis buffer for 1 h on ice and then incubated with rabbit anti-HA mAb-coupled protein A/G beads overnight at 4 °C. Subsequently, the beads were centrifuged at 2000 × g for 5 min at 4 °C and washed three times with PBS. Finally, the bead-bound proteins were eluted with lysis buffer and analyzed by Western blot analysis.

Indirect immunofluorescence assay (IFA)

HEK293T cells were fixed in 3.7% paraformaldehyde in PBS at room temperature for 30 min, followed by 0.1 M glycine in PBS for 10 min. The cells were subsequently washed with PBS 3 times, permeabilized with 0.3% Triton X-100 at 4 °C for 15 min, and blocked with 1% bovine serum albumin (BSA) at room temperature for 1 h. The cells were then incubated with the indicated primary antibodies at room temperature for 1 h. After being washed as previously described, the cells were incubated with the secondary antibody for 1 h at room temperature. The primary antibodies used were mouse anti-HA tag mAb (1:1,000 dilution; Sigma–Aldrich, USA), rabbit anti-Flag tag mAb (1:1,000 dilution; Sigma–Aldrich, USA), rabbit anti-6×His tag mAb (1:1,000 dilution; Proteintech, USA), rabbit anti-SPOP pAb (1:500 dilution; Thermo Fisher Scientific, USA), rabbit anti-p62 pAb (1:500 dilution; Sigma–Aldrich, USA), and anti-PRRSV N protein mAb (1:500 dilution; prepared in this laboratory) [47]. The secondary antibodies, which were diluted 1:1000 in PBS, were Alexa Fluor 488-conjugated goat anti-mouse IgG (H + L) and Alexa Fluor 568-conjugated goat anti-rabbit IgG (H + L) (Thermo Fisher Scientific, USA). The cells were counterstained with DAPI and embedded in Prolong Gold (Thermo Fisher Scientific, USA) for further analysis. The samples were analyzed with a Fast Airyscan confocal laser scanning microscope (LSM980-ZEISS, USA).

Bimolecular fluorescence complementation (BiFC) assay

The construction of the BiFC plasmids was based on the splitting of the YFP fluorescent protein into two inactive fragments, each of which was ligated to the target protein and expressed intracellularly in a fusion. In this study, the BiFC system was used to confirm the intercellular interaction between Nsp9 and 9nb, 9nb-SPOP, 9nb-SPOP_ΔNLS, 9nb-SPOP_mutΔNLS or NCnb-SPOP in HEK293T cells. Briefly, the 9nb, 9nb-SPOP, 9nb-SPOP_ΔNLS, 9nb-SPOP_mutΔNLS and NCnb-SPOP sequences were amplified and inserted into the pVC plasmid, and Nsp9 was inserted into the pVN plasmid via homologous recombination as described previously [48]. The constructed plasmids were named pVC-9nb, pVC-9nb-SPOP, pVC-9nb-SPOP_ΔNLS, pVC-9nb-SPOP_mutΔNLS, pVC-NCnb-SPOP and pVN-Nsp9. HEK293T cells were cotransfected with pVN-Nsp9 and pVC-9nb, pVC-9nb-SPOP, pVC-9nb-SPOP_ΔNLS, or pVC-NCnb-SPOP at a 1 to 1 ratio (pVC and pVN containing only the ORF of VC or VN were used as controls). After 24 hpt, the cells were incubated with NucBlue™ Live Cell Stain ReadyProbes™ (Thermo Fisher Scientific, USA) for 10 min at 37 °C while protected from light, and live-cell imaging was performed with a Fast Airyscan confocal laser scanning microscope (LSM980-ZEISS, USA). The primer sequences are shown in Table 1.

Expression and purification of prokaryotic protein

The pET-28a-His-Nsp9, pET-28a-His-Nsp9-mCherry, pET-28a-His-N-mCherry, pGEX-6P-1-FLAG-pre-9nb-SPOP, pGEX-6P-1-FLAG-9nb-SPOP, pGEX-6P-1-FLAG-RNF4-RING and pGEX-6P-1-FLAG-NCnb-SPOP recombinant plasmids were transformed into E. coli Rosetta (DE3) competent cells. Protein expression was induced with 0.2 mM IPTG when the OD₆₀₀ reached 0.6, followed by incubation at 16℃ for 18 h. The bacterial cells were harvested by centrifugation at 6000×g for 10 min and resuspended in binding buffer (50 mM Tris-HCl, 100 mM NaCl and 5% glycerol, pH 8.0). After sonication and centrifugation, the supernatant was then filtered through a 0.22‐µm filter. His-tagged recombinant proteins (His-Nsp9, His-Nsp9-mCherry, and His-N-mCherry) were loaded into a gravity‐flow column containing Ni‐NTA chelated agarose. Target proteins were eluted using an elution buffer (50 mM Tris‐HCl, 100 mM NaCl, 5% glycerol and 200 mM imidazole, pH 8.0) [49]. Purification of the soluble recombinant proteins (GST-pre-9nb-SPOP, GST-9nb-SPOP, and GST-NCnb-SPOP) was performed with a GST-tag Protein Purification Kit (Beyotime Biotechnology, China) following the manufacturer’s protocol. The collected proteins were subsequently analyzed by Coomassie Brilliant Blue staining or Western blotting. The purified protein concentrations were measured by a BCA protein concentration determination kit (Thermo Fisher Scientific, USA), and the proteins were stored at -80℃ for future use.

Indirect enzyme-linked immunosorbent assay (ELISA)

To determine the binding activity of pre-9nb-SPOP, 9nb-SPOP and NCnb-SPOP (negative control) with Nsp9, a modified ELISA was designed and performed as previously described [30]. Purified His-Nsp9 protein was diluted to 10 µg/mL in carbonate coating buffer (pH 9.6) and coated onto an ELISA plate (Biofil, China) with 100 µL/well overnight at 4℃. After washing with PBST three times, the plate was blocked with 200 µL of 5% skim milk (BD, USA) in PBST for 1 h at 37℃. Following three washes with PBST, the plate was dried. Separately, a series of eight two-fold dilutions (10–1280 mM) were prepared for each protein: GST-pre-9nb-SPOP, GST-9nb-SPOP, and GST-NCnb-SPOP. Then, 100 µL of each dilution was added to the plate as the first antibody and incubated for 1 h at 37℃. After washing again, mouse anti-GST monoclonal antibody (1:1000, Abcam, England) was added and incubated for 1 h at 37℃. Following incubation, 100 µL of HRP-conjugated goat anti-mouse IgG (1:10,000, Proteintech, USA) was added per well. After a third washing, the single-component substrate solution tetramethylbenzidine (TMB) (Solarbio, China) was added to produce a color reaction. Finally, 2 M H₂SO₄ was used to stop the reaction, and OD values were read at 450 nm by an automated ELISA plate reader.

Establishment of stable Marc-145 cell lines expressing bioPROTACs

When the cells reached 80%–90% confluence, HEK293T cells were transfected with a plasmid transfer vector (PLEX-9nb-SPOP-IRES-mCherry or PLEX-9nb-SPOP_ΔNLS-IRES-mCherry) and the packaging plasmids psPAX2 and pMD2.G at a 3:2:1 ratio with X-tremeGENE HP DNA Transfection Reagent (Roche Applied Science, Penzberg, Germany) according to the manufacturer’s instructions. The lentivirus was harvested at 24 and 48 hpi and stored at 4 °C. Marc-145 cells (1.0 × 10⁵) were evenly inoculated in 6-well plates and then incubated with mixtures of 1 mL of DMEM and 1 mL of lentivirus at 37 °C for 1.5–2 h when the cell density reached 50%. At 48 h after lentiviral infection, the positive cells expressing mCherry were sorted into 96-well plates at a density of one cell per well via a SONY-MA900 Flow Cell Sorter. The sorted monoclonal cells were expanded for culture and monitored by fluorescence microscopy. To ensure the purity of the positive cell population remained above 95%, flow cell sorting was performed after every five passages. A cell viability assay was performed with an MTT Cell Proliferation and Cytotoxicity Assay Kit (Beyotime, China) according to the manufacturer’s instructions. Cell viability was defined as the percentage of absorbance compared with that of the control at OD570.

Virus titration

Marc-145 cells were seeded on 96-well plates for 24 h prior to viral inoculation and then incubated with 100 µL of viral supernatant in tenfold serial dilutions. Each virus dilution consisted of 8 replicates. Five days after inoculation, the 50% tissue culture infectious dose (TCID₅₀) was calculated using the Reed–Muench method.

Flow cytometry analysis

After 24 h of transfection or drug treatment, 10⁶ cells were collected, incubated with 1 mL of fixation/permeabilization working solution for 60 min at 4 °C, and then blocked with 1% BSA for 1 h at 37 °C. The samples were incubated with a monoclonal anti-HA antibody (1:1,000 dilution) or PRRSV N monoclonal antibody for 1 h at 37 °C. After centrifugation and washing with PBS three times, an Alexa Fluor 488-conjugated goat anti-mouse secondary antibody (1:800 dilution) was added to the cells, and the cells were incubated for 1 h at room temperature. The cells were collected and resuspended in flow cytometry staining buffer, after which the samples were analyzed on a BD LSRFortessa X-20.

Reverse transcription–qPCR (RT–qPCR)

The viral RNA of cells, tissue, serum, and swab samples was extracted with a TIANamp Virus RNA Kit (Tiangen, China) and subjected to RT–qPCR to quantify the copy number of PRRSV RNA via a One Step PrimeScript™ RT–PCR Kit (TaKaRa, China) according to the manufacturer’s instructions. The primers and probe sequences of PRRSV-1 [50] and PRRSV-2 [51] used for qPCR amplification are listed in Table 1.

Serial passages of the PRRSV HuN4 strain under the selection of 9nb-SPOP_ΔNLS

To study the emergence of bioPROTAC-resistant PRRSV variants (designated HuN4-bioPROTAC), we used the Marc-145 cell line stably expressing bioPROTAC (Marc-145-9nb-SPOP_ΔNLS, expression rate greater than 90%). This system provided sustained high‑level expression of bioPROTACs and consistent drug selection pressure during serial passaging of the PRRSV HuN4 strain. Previous studies have reported that ribavirin is effective against various RNA viruses, including PRRSV, foot-and-mouth disease virus (FMDV), and hepatitis C virus (HCV) [52]. Serial passaging of PRRSV in Marc-145 cells containing 0.1 and 0.2 mM ribavirin was able to induce the emergence of ribavirin-resistant mutant viruses [52, 53]. It has been reported that the in vitro EC₅₀ of ribavirin against PRRSV infection was 23.7 ± 3.1 µM [54], and 10× EC₅₀ was wildly used to select for drug-resistant variants [55]. Therefore, we selected 0.2 mM ribavirin as the pharmacological control to induce ribavirin-resistant PRRSV variants (HuN4-RBV) in this study. Briefly, confluent monolayer of Marc-145 cells was prepared in 6-well plates and pretreated overnight at 37℃ with DMEM medium containing ribavirin to simulate the infection conditions of the stable 9nb-SPOP_ΔNLS expressing cell line. The confluent Marc-145-9nb-SPOP_ΔNLS or ribavirin-pretreated Marc-145 cells were infected by the HuN4 strain at an MOI of 0.1. After 1 h of incubation, the viral inoculum was removed, and fresh DMEM medium or medium containing 0.2 mM ribavirin was added, respectively. The infection was then allowed for additional 48 h, after which the supernatant was collected from each well, centrifuged, and stored at -80℃ until use. A total of 500 µL of the supernatant from each passage was used as the inoculum for the next passage. This procedure was repeated a total of 15 times. The supernatant was collected for Nsp9 gene sequencing, and the viral titer was determined in Marc-145 cells.

Assessment of the inhibitory effect of 9nb-SPOP_ΔNLS on the ribavirin-resistant mutant

The replication of the ribavirin-resistance variant (HuN4-RBV-P15) was assessed in Marc-145 cells with or without the presence of 9nb-SPOP_ΔNLS. Confluent monolayers of Marc-145-9nb-SPOP_ΔNLS cells were prepared in 6-well plate, inoculated with RBV-P15 at an MOI of 0.01, and incubated for 1 h. After incubation, the non-attached virus was removed, and fresh DMED medium was added. The infected cells were then incubated for 72 h. The supernatants were collected for virus titration in Marc-145 cells, and the cells were lysed for quantification of viral load by RT–qPCR.

Degradation effects of bioPROTACs in vivo

To evaluate the in vivo degradation efficacy of bioPROTACs on target proteins, eukaryotic vectors expressing bioPROTAC proteins (9nb-SPOP_ΔNLS and Nnb-SPOP) and target proteins (PRRSV Nsp9 and N protein) fused to red fluorescent protein-mCherry were constructed. These constructs were codelivered via hydrodynamic tail vein injection using the TransIT-EE (Enhanced) transfection system using a volume calculated on the basis of mouse body weight: injection volume (mL) = [body weight (g)/10] + 0.1 mL. Plasmids and the required volume of injection solution were added to sterile EP tubes before injection. The samples were mixed and allowed to incubate for 10 min before being injected into 6-week-old BALB/c mice at a constant rate. The animals were sacrificed 24 h postinjection, a time point at which high transgene expression was confirmed in major organs. Fluorescence imaging was performed using a small animal in vivo imaging system (Berthold NightOWL LB 983, Germany) under anesthesia. The mCherry signal was excited at 500 nm, and the emission was measured at 600 nm. The total fluorescence intensity within regions of interest was quantified using IndiGO 2 software. To validate the expression and degradation of the target protein, lung tissues were homogenized and subjected to western blot analysis.

Production of bioPROTACs fused to the porcine IgG Fc fragment

On the basis of the anti-PRRSV activity of bioPROTACs in stably transfected Marc-145 cells, we sought to extend these in vitro findings to in vivo applications. To enable targeted delivery to porcine alveolar macrophages (PAMs), the primary host cells of PRRSV, we fused each bioPROTAC construct to a porcine IgG Fc fragment. The gene encoding the porcine Fc fragment (GenBank: AK405781) was synthesized by Ruibiotech Co., Ltd. (China). This fragment was subsequently cloned and inserted into prokaryotic expression vectors encoding 9nb-SPOP_ΔNLS, Nnb-SPOP, NCnb-SPOP, and two nanobody genes by amplification primers (Table 1). All the constructs were verified by DNA sequencing. The confirmed plasmids were transformed into Rosetta (DE3) cells for recombinant protein expression. BioPROTAC-Fc fusion proteins were expressed by induction with 0.1 mM IPTG at 16 °C for 20 h. Bacterial cells were harvested and lysed by ultrasonication, and the soluble fraction was purified via nickel-affinity chromatography following established protocols.

Analysis of bioPROTAC-Fc entering PAMs

The PAMs (1 × 10⁶ cells/mL) were cultured in 6-well plates for 6 h and then incubated with 9nb-SPOP_ΔNLS-Fc, Nnb-SPOP-Fc, NCnb-SPOP-Fc, 9nb-Fc, or Nnb-Fc at final concentrations ranging from 0 to 50 µg/mL for 2 h. After treatment, the cells were collected or fixed and analyzed by Western blotting or IFA.

BioPROTACs with the pFc fusion protein inhibit PRRSV replication in PAMs

PAMs were seeded in 6- or 24-well plates at a density of 1 × 10⁶ cells/mL and cultured for 4 h, followed by treatment with bioPROTAC-Fc fusion proteins at a final concentration of 10 µg/mL for 2 h. Next, the cells were infected with HP-PRRSV (HuN4) at 0.01 MOI. At 24 hpi, the cells were collected or fixed and analyzed by Western blotting and IFA. The culture supernatants were collected for progeny virus titration in Marc-145 cells, and the cells were lysed for quantification of the viral load by RT–qPCR.

Preparation of bioPROTAC-mRNA-LNP

To construct pre-linearized template vector (LTV) for bioPROTAC-mRNA-LNP synthesis, a Cloning Kit for mRNA Template (TaKaRa, China) was used. LTV-Nnb-SPOP and LTV-9nb-SPOP_ΔNLS were generated by subcloning the respective sequences into the LTV plasmid using specific primers (Fig. S7A). IFA confirmed the successful expression of plasmid-encoded Nnb-SPOP and 9nb-SPOP_ΔNLS in HEK293T cells, supporting their suitability for subsequent in vitro transcription and LNP packaging procedures (Fig. S7B). Each plasmid was then linearized with the restriction enzyme Hind III (300 µg per reaction). The digested products were purified by extraction with an equal volume of phenol: chloroform: isoamyl alcohol (25:24:1, pH = 8.0), followed by vortexing and centrifugation at 12,000 × g for 2 min. The aqueous phase was collected, and nucleic acids were precipitated using 0.1 volume of 3 M sodium acetate (pH = 5.2) and an equal volume of isopropyl alcohol, incubated on ice for 5 min, and centrifuged at 12,000 × g for 10 min at 4 °C. The pellet was subsequently washed with 1 mL of 70% ethanol, centrifuged again at 12,000 × g for 5 min, air-dried, and resuspended in DEPC-treated water. Furthermore, mRNA was synthesized by in vitro transcription from linearized LTV-Nnb-SPOP and LTV-9nb-SPOP_ΔNLS plasmids. The resulting mRNA products were then encapsulated into lipid nanoparticles (LNPs) at a 1:1 ratio to form the final product, designated mRNA (N + 9). mRNA synthesis and LNP encapsulation were performed by CSPC PHARMA, Inc. All primers were synthesized by Ruibo Xingke Biotechnology, Inc., and their sequences are listed in Table 1.

Antiviral effect of bioPROTACs in piglets

Twenty 4-week-old healthy Landrace pigs (6–8 kg) were obtained from a commercial farm in Heilongjiang Province. Prior to the study, all the pigs were confirmed to be negative for antigens or antibodies against African Swine Fever Virus (ASFV), Classical Swine Fever Virus (CSFV), and PRRSV via RT–PCR and commercial ELISA kits. All the pigs were housed in the BSL-2 (Biosafety Level 2) facilities of the research institute, with oversight provided by professional veterinarians. To ensure animal welfare, humane endpoints were strictly observed. Any pig showing an inability to eat or stand independently, along with a significantly diminished response to environmental stimuli, was euthanized promptly to alleviate suffering.

To evaluate the antiviral efficacy of bioPROTACs in vivo, piglets were randomly divided into five groups (n = 4 per group), with each group housed in an independent isolation room. Given that the in vivo delivery efficiency and effective concentration maintenance of bio-macromolecular drugs limit their antiviral efficacy, we administered the bioPROTAC 24 h prior to viral challenge in piglets to ensure that effective drug levels were reached at viral infection. This animal experimental protocol is a modification of the previously published protocol for antiviral animal studies [56, 57]. The detailed experimental grouping is presented in Table 2. Following a two-day acclimatization period, all piglets except those in the negative control group were challenged intramuscularly with 1 × 10⁵ TCID₅₀ of the HP-PRRSV HuN4 strain. The negative control group received an equivalent volume of PBS via the same route. At -1, 1, and 3 days postinfection (dpi), the piglets in the corresponding drug treatment groups were administered 1 mg of Protein (N + 9), Protein (NCnb), or mRNA (N + 9) (concentration: 1 mg/mL) via auricular vein injection. The negative control and challenge control (Mock) groups received an equal volume of PBS following the same injection procedure.

Table 2.

Animal groups and corresponding treatments

Group name	Injected bioPROTACs	Challenge PRRSV
Negative control	PBS	PBS
Mock/HP-PRRSV	PBS	HuN4
Protein(N + 9)/HP-PRRSV	Protein(N + 9)	HuN4
Protein(NCnb)/HP-PRRSV	Protein(NCnb)	HuN4
mRNA(N + 9)/HP-PRRSV	mRNA(N + 9)	HuN4

Clinical symptoms were monitored, and rectal temperature was measured daily. Clinical scores were assessed and recorded at 0 and 7 dpi. All the piglets were euthanized at 10 dpi for necropsy. A macroscopic examination of pulmonary pathological changes was first performed, after which lung tissue samples were collected for histopathological examination. The extent of gross lung lesions was evaluated based on a previously established standard scoring system [29]. All lung tissue samples were fixed in 10% neutral buffered formalin and embedded in paraffin. Sections were prepared and stained with hematoxylin and eosin (H&E) for histopathological examination under microscopy. Additionally, various tissues were collected as described above for the quantification of viral copy number. To evaluate viremia and viral shedding during the early phase of infection, blood samples, anal swabs, and nasal swabs were collected from the pigs in each group at 1 and 4 dpi. The viral load in these samples was quantified using the above-described methods.

Bioinformatics analysis

The spatial structures of 9nb-SPOP and 9nb-RNF4 were predicted with the online I-TASSER server (http://zhanglab.dcmb.med.umich.edu/I-TASSER). The structure files of 9nb-SPOP and 9nb-RNF4 were analyzed using PyMOL 2.5.5 software.

Statistical analysis

All experiments were performed with three independent replicates, and the error bars indicate the standard deviation (SD). Statistically significant differences were analyzed using the two-tailed Student’s t test for comparisons between two groups and one-way analysis of variance (ANOVA) with the Tukey multiple-comparison test or two-way ANOVA with Dunnett’s multiple comparison test for comparisons among multiple groups (GraphPad Prism 8.0 software). Survival curves were analyzed for P values using the log-rank (Mantel–Cox) test. A P value < 0.05 was considered to indicate statistical significance. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.5; ns, P > 0.05.

Results

SPOP-derived degraders mediate targeted degradation of PRRSV viral proteins

PRRSV Nsp9 protein localizes in both the nucleus and cytoplasm during infection, therefore, we selected the representative nuclear CRL3 ligase adaptor SPOP for degrader design. SPOP can efficiently recruit and bind to Cullin 3 through its conserved BTB domain, thereby assembling into an active complex [58]. SPOP consists of the natural substrate recognition domain MATH, a BTB structural domain, a BACK domain and an NLS (Fig. 1A). The efficiency of target protein degradation has been shown to correlate with the substrate-binding affinity of the bioPROTAC ligand [11]. Given that dual nanobodies exhibit higher binding affinity than that of a single nanobody [59], we designed a bioPROTAC molecule targeting the PRRSV Nsp9 protein by replacing the MATH domain of SPOP with an Nsp9-specific nanobody (Nsp9nb). To further enhance the binding affinity, two Nsp9nbs units were linked with a short flexible linker (GGGGS)2 and subsequently fused to the N-terminus of the SPOP BTB domain (Fig. S1A), generating the fusion protein designated as 9nb-SPOP. We initially assessed the binding activity of 9nb-SPOP to Nsp9 by Co-IP, and Nsp9 could engage with 9nb-SPOP and its NLS-deficient mutant (9nb-SPOP_ΔNLS). Nsp9 was specifically coimmunoprecipitated with an anti-flag antibody, confirming successful binding (Fig. 1B). In addition, the affinity of 9nb-SPOP, single Nsp9nb-fused SPOP (pre-9nb-SPOP), and nonspecific nanobody-fused SPOP (NCnb-SPOP) with Nsp9 was analyzed by ELISA using a purified Nsp9 protein-coated plate. The results revealed that 9nb-SPOP had the highest binding affinity to Nsp9 in a dose-dependent manner. Notably, at 1280 µM, the affinity of 9nb-SPOP was threefold greater than that of pre-9nb-SPOP (Fig. 1C). Furthermore, we performed a bimolecular fluorescence complementation (BiFC) assay, which has been extensively applied to study protein–protein interactions [48, 60], to confirm specific interactions. HEK293T cells were cotransfected with the indicated plasmids, including pVC-9nb, pVC-9nb-SPOP, pVC-NCnb-SPOP, or pVC-control, and pVN-Nsp9. As expected, Nsp9 engaged well with 9nb and 9nb-SPOP but did not interact with NCnb-SPOP or the pVC empty vector control (Fig. 1D). The Nsp9-9nb complex was distributed mainly in the cytoplasm. However, the Nsp9-9nb-SPOP complex was detected primarily in the nucleus, exhibiting the typical nuclear dimerization and multimerization properties of SPOP [58], and only a few complexes were present in the cytoplasm (Fig. 1D). When the pVN empty vector was cotransfected with the above plasmids, no interactions were detected (Fig. S3A).

Fig. 1.

Nanobody-mediated bioPROTACs induce targeted degradation of PRRSV Nsp9. (A) Schematic diagram of the SPOP protein domain. The SPOP protein is composed of the MATH domain (aa 28–166), BTB domain (aa 28–166), BACK domain (aa 300–359), and a C-terminal nuclear localization signal (NLS). (B) Coimmunoprecipitation of Nsp9 with 9nb-SPOP and 9nb-SPOP_ΔNLS. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9 or control plasmids with pcDNA3.1-Flag-9nb-SPOP or pcDNA3.1-Flag-9nb-SPOP_ΔNLS. After 24 h, the cells were subjected to IP with an antibody against Flag, followed by Western blot analysis. (C) The binding activity of pre-9nb-SPOP, 9nb-SPOP and NCnb-SPOP (negative control) with Nsp9 by iELISA. The data are presented as the means ± SDs. (D) BiFC analysis for 9nb-SPOP interactions with Nsp9. HEK293T cells were cotransfected with 1 µg of pVN-Nsp9 and pVC-9nb, pVC-9nb-SPOP or pVC-NCnb-SPOP. pVC and pVN were used as empty control plasmids. After 24 h, the nuclei were stained with NucBlue™ Live ReadyProbes™, and live images were acquired under a 100× lens objective for confocal microscopy. The interacting proteins are shown in yellow. Scale bars, 10 μm. (E) IFA analysis of the degradation efficiency of 9nb-SPOP and 9nb-RNF4. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 3 µg control plasmids, pcDNA3.1-Flag-9nb-SPOP or pcDNA3.1-Flag-9nb-RNF4-RING. After 24 h, the cells were fixed, permeabilized, and subjected to anti-HA antibody (green). Fluorescent signals were visualized by confocal microscopy. Scale bars, 100 μm. (F) The fluorescence of (E) was measured by ImageJ software. The statistical results of the fluorescence percentage are presented as a column chart from three independent experiments. (G) Western blot analysis of 9nb-SPOP degradation efficiency. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 0, 1, 2, or 3 µg pcDNA3.1-Flag-9nb-SPOP or control plasmids for 24 h. Band intensities from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3). (H) Flow cytometry analyses of the degradation efficiency of Nsp9 by 9nb-SPOP at different doses. (I) The fluorescence percentage of (H) is presented as a column chart from three independent experiments. All the experiments were performed three times, and the error bars indicate the SDs. Statistical significance was determined using the Student’s t test; ***, P < 0.001; **, P < 0.01; ns, P > 0.05

To confirm that the observed degradation was specifically mediated by the recruitment of the SPOP E3 ubiquitin ligase adaptor, we included a control construct in which Nsp9nb was fused to the N-terminal SUMO substrate binding site of RNF4 (9nb-RNF4; Fig. S1B-C), which is an E3 ligase [61]. The degradation of Nsp9 in HEK293T cells was analyzed by IFA (Fig. 1E). Compared with cells expressing Nsp9 alone, those cotransfected with 9nb-SPOP exhibited a notable reduction in Nsp9 expression. Quantitative analysis confirmed that the percentage of green fluorescent cells decreased from approximately 23% to 3% following 9nb-SPOP treatment (P < 0.001, Fig. 1F). Although GST pull-down and Co-IP assays confirmed an interaction between 9nb-RNF4 and Nsp9 (Fig. S1D and E), Nsp9 expression was not significantly decreased in in cells cotransfected with 9nb-RNF4 (P > 0.05; Fig. 1E and F). Consistent with this observation, Western blot analysis further demonstrated that 9nb-RNF4 had no appreciable effect on Nsp9 protein levels (Fig. S1F). In contrast, 9nb-SPOP promoted the degradation of Nsp9 in a dose-dependent manner. Co-transfection with 3 µg of 9nb-SPOP resulted in approximately 60% reduction of Nsp9 protein in HEK293T cells (Fig. 1G; P < 0.001). Moreover, flow cytometry analyses demonstrated that following cotransfection with 1 µg and 3 µg of the 9nb-SPOP plasmid, the percentage of cells expressing Nsp9 significantly decreased from 23% to 10% and 6%, respectively (P < 0.001; Fig. 1H and I).

To further evaluate the general applicability of the SPOP-derived bioPROTAC platform, the PRRSV N protein was selected as a degradation target protein due to its low amino acid mutation rate in both PRRSV-1 and PRRSV-2 subtypes. Based on our previous study that indicated that nanobody Nb1 has high affinity for the N protein [46], we engineered a new bioPROTAC construct by substituting the Nsp9nb with a Nb1, named Nnb-SPOP (Figure S2A). Western blot analysis confirmed that N protein levels were reduced in a dose-dependent manner with Nnb-SPOP cotransfection, demonstrating a maximum reduction exceeding 50% (Figure S2B). These results demonstrated the successful establishment of a modular bioPROTAC platform comprising nanobodies, SPOP, and flexible linkers, which enabled the targeted degradation of multiple viral antigens simply by exchanging target-specific nanobodies.

Depletion of the ubiquitination function of SPOP inhibits the degrader activity of bioPROTAC

To confirm that the degradation of 9nb-SPOP is dependent on the ubiquitination machinery of SPOP, as the three-box motif significantly affects the recruiting CUL3 function of SPOP [11], control constructs were engineered carrying a deletion of the 302–326 residues of SPOP (9nb-SPOP_mut). As expected, Co-IP assays revealed that ubiquitinated Nsp9 was readily detected in HEK293T cells cotransfected with 9nb-SPOP or the 9nb-SPOP_ΔNLS mutant, confirming that 9nb-SPOP induced Nsp9 ubiquitination and that the lack of a nuclear localization signal did not affect its ubiquitination function (Fig. 2A). Rather, no ubiquitination signal was detected in cells expressing 9nb-SPOP_mut, indicating that 9nb-SPOP_mut lost its ability to induce nsp9 ubiquitination (Fig. 2B). Moreover, an obviously decreased expression level of Nsp9 was detected in 9nb-SPOP-transfected cells but not in 9nb-SPOP_mut control-transfected cells by confocal microscopy (Fig. 2C), suggesting that 9nb-SPOP_mut did not affect Nsp9 depletion. Although the colocalization of Nsp9 and 9nb-SPOP_mut was still observed in the nucleus, as indicated by the white arrows, it lacked the typical liquid-droplet-like multimerization properties of SPOP (Fig. 2C). These results were consistent with those of the Western blot analysis, which showed that 9nb-SPOPmut was unable to induce degradation of the Nsp9 protein (P > 0.05; Fig. 2D), while 9nb-SPOP (WT) effectively induced Nsp9 degradation (Fig. 1G). These results indicated that the ubiquitination function of the 3-box domain was crucial for 9nb-SPOP degradation activity.

Fig. 2.

Degradation of Nsp9 by 9nb-SPOP is dependent on the ubiquitination machinery of SPOP. (A) Coimmunoprecipitation analysis of Nsp9 ubiquitination mediated by 9nb-SPOP and 9nb-SPOP_ΔNLS. (B) Coimmunoprecipitation analysis of Nsp9 ubiquitination by the 9nb-SPOP mutant. (C) IFA of the degradation of Nsp9 by 9nb-SPOP and 9nb-SPOP_mut. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 3 µg control plasmids, pcDNA3.1-Flag-9nb-SPOP or pcDNA3.1-Flag-9nb-SPOP_mut. After 24 h, the cells were fixed, permeabilized, and subjected to anti-HA antibody (green) and anti-Flag antibody (red). Nuclei (blue) were labeled using DAPI. The colocalization of Nsp9 and 9nb-SPOP_mut is indicated by the arrow. Scale bars, 50 μm. (D) Western blot analysis of 9nb-SPOP_mut degradation efficiency. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 0, 1, 2, or 3 µg pcDNA3.1-Flag-9nb-SPOP_mut or control plasmids in 6-well plates for 24 h. Band intensities from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3). All the experiments were performed three times, and the error bars indicate the SDs. Statistical significance was determined using the Student’s t test; ns, P > 0.05

9nb-SPOP enables targeted degradation of Nsp9 in the cytoplasm

The pathway involved in the SPOP-derived PROTAC-mediated downregulation of PRRSV Nsp9 and N protein expression was investigated. First, HEK293T cells cotransfected with 9nb-SPOP and Nsp9 were treated with MG132 and chloroquine (CQ), a proteasome inhibitor and a lysosomal autophagy inhibitor, respectively. Interestingly, the depletion of Nsp9 was partially inhibited by both MG132 and CQ, and the inhibitory activity of MG132 was greater than that of CQ (Fig. 3A). These results suggested that Nsp9 degradation induced by 9nb-SPOP occurred through both the proteasome and the lysosome, involving two major cellular clearance pathways. However, MG132 markedly inhibited degradation of the N protein, while CQ had no inhibitory effect (Fig. 3B). N protein has been reported to mainly accumulate in the cell nucleus, with a smaller amount in the cytoplasm, which allows free distribution between the cell nucleus and cytoplasm [62]. Thereby, we speculate that Nnb-SPOP mainly degrades the nuclear N protein through the ubiquitin-proteasome pathway, but the degradation of the cytoplasmic N protein may involve other mechanisms. On the basis of these findings, we propose that the degradation pathway employed by bioPROTACs may depend on the subcellular localization of the target protein. We first verified that removing the NLS from SPOP preserved its ubiquitination of SPOP substrates (Fig. 2A), which was consistent with the findings of a previous study [58]. Next, to confirm our hypothesis, we examined whether SPOP lacking the NLS would accumulate in the cytoplasm and eliminate cytoplasmic Nsp9. The intracellular distribution of 9nb-SPOP_ΔNLS and Nsp9 was examined by confocal microscopy. As expected, 9nb-SPOP_ΔNLS was expressed in the cytoplasm and was not localized to the nucleus (Fig. 3C). In addition, when 9nb-SPOP_ΔNLS and Nsp9 were cotransfected into HEK293T cells, Nsp9 colocalized with 9nb-SPOP_ΔNLS, and liquid-droplet-like multimerization was observed in the cytoplasm, as indicated by the white arrows (Fig. 3C). These findings confirmed that cytoplasmically localized 9nb-SPOP_ΔNLS colocalized with Nsp9 and induced the formation of liquid-droplet-like multimerization, confirming the successful formation of the ternary complex.

Fig. 3.

9nb-SPOP mediates Nsp9 degradation in the cytoplasm. BioPROTAC-induced degradation of the Nsp9 protein (A) and N protein (B) and rescue by proteasomal or lysosomal inhibition. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9 (A) or pCAGGS-HA-N (B) together with the indicated amounts of pcDNA3.1-Flag-9nb-SPOP (A) or pcDNA3.1-Flag-Nnb-SPOP (B) and control plasmids for 24 h, followed by treatment with 10 µM MG132, 50 µM chloroquine (CQ), or DMSO for 6 h. Protein degradation was analyzed by Western blot analysis. Band intensities from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3). (C) IFA detection of the colocalization of Nsp9 and 9nb-SPOP_ΔNLS. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 3 µg of pcDNA3.1-Flag-9nb-SPOP or pcDNA3.1-Flag-9nb-SPOP_ΔNLS. After 24 h, the distribution of Nsp9 (green) and bioPROTAC (red) was detected. Representative colocalization signals are indicated by white arrows. Scale bars, 50 μm. (D) BiFC analysis of the interaction between 9nb-SPOP_ΔNLS and Nsp9 in the cytoplasm. HEK293T cells were cotransfected with 1 µg pVN-Nsp9 and 1 µg pVC-9nb-SPOP, pVC-9nb-SPOP_ΔNLS or pVC-9nb-SPOP_mut△NLS. pVC was used as an empty control plasmid. After 24 h, the nuclei were stained with NucBlue™ Live ReadyProbes™, and live images were acquired under a 100× lens objective for confocal microscopy. The interacting proteins are shown in yellow. Scale bars, 10 μm. (E) Western blot analyses of 9nb-SPOP_ΔNLS degradation efficiency. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 0, 1, 2, or 3 µg pcDNA3.1-Flag-9nb-SPOP_ΔNLS or control plasmid for 24 h. Band intensities from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3). (F) IFA results for the degradation efficiency of 9nb-SPOP_ΔNLS and 9nb-SPOP_mutΔNLS. HEK293T cells were cotransfected with 1 µg pCAGGS-HA-Nsp9 and 3 µg control plasmids, pcDNA3.1-Flag-9nb-SPOP_ΔNLS or pcDNA3.1-Flag-9nb-SPOP_mutΔNLS. After 24 h, the expression of Nsp9 (green) was detected. Scale bars, 100 μm. (G) The fluorescence of (F) was measured by ImageJ software. The statistical results of the fluorescence percentage are presented as a column chart from three independent experiments. All the experiments were performed three times, and the error bars indicate the SDs. Statistical significance was determined using the Student’s t test; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns, P > 0.05

Interestingly, when the interactions between Nsp9 and 9nb-SPOP_ΔNLS were analyzed via a BiFC assay, only a small number of aggregations were induced by 9nb-SPOP_ΔNLS in the cytoplasm, while a typical multimerized formation in the nucleus was induced by 9nb-SPOP (Fig. 3D). When the pVN empty vector was cotransfected with the above plasmids, no interactions were detected (Fig. S3B). To distinguish between impaired binding and actual protein degradation as the cause of the low Nsp9-9nb-SPOP_ΔNLS complex signal, we generated a ubiquitination-deficient mutant of 9nb-SPOP_ΔNLS (9nb-SPOP_mutΔNLS). This mutant was designed to bind Nsp9 but was unable to catalyze its ubiquitination. The 9nb-SPOP_mutΔNLS retained its ability to bind Nsp9 predominantly in the cytoplasm. Notably, the level of the stable ternary complex formed by the catalytically inactive mutant was substantially greater than that formed by the functional 9nb-SPOP and 9nb-SPOP_ΔNLS degraders (Fig. 3D). These results demonstrated that the weak signal of the Nsp9-9nb-SPOP_ΔNLS complex was a consequence of Nsp9 degradation and that ubiquitination was critical for the cytoplasmic degradation of Nsp9. Furthermore, 9nb-SPOP_ΔNLS-mediated Nsp9 degradation was confirmed at the protein level, and 60% of Nsp9 was depleted by 3 µg of 9nb-SPOP_ΔNLS (P < 0.001; Fig. 3E). Moreover, when Nsp9 was cotransfected with 9nb-SPOP_ΔNLS, the Nsp9 expression level decreased compared with that in the Nsp9 group and 9nb-SPOP_mutΔNLS-cotransfected groups (Fig. 3F). A quantification of the percentage of fluorescence revealed that Nsp9 expression was approximately 50% lower in the cells cotransfected with 9nb-SPOP_ΔNLS than in the Nsp9 control cells (P < 0.001; Fig. 3G). These results indicated that although SPOP was a nucleus-localized E3 ubiquitin ligase ligand, nanobody-fused SPOP could mediate functional cytoplasmic degradation of its ubiquitinated substrate.

9nb-SPOP_ΔNLS mediates Nsp9 cytoplasmic degradation through p62-mediated selective autophagy

To clarify the protein degradation pathway activated by 9nb-SPOP_ΔNLS, HA-Nsp9 and Flag-9nb-SPOP_ΔNLS-cotransfected cells were treated with MG132 or CQ. With CQ treatment, the depletion of Nsp9 was blocked, whereas treatment with MG132 resulted in less inhibition of Nsp9 degradation (Fig. 4A). Moreover, the Nsp9 expression level was analyzed by flow cytometry after treatment with MG132 and CQ. The results showed that CQ treatment significantly inhibited the depletion of Nsp9 by either 1–3 µg of 9nb-SPOP_ΔNLS, and the percentages of Nsp9-expressing cells reached 30% and 25%, respectively, which were 24% and 17%, respectively, in the DMSO group (Fig. 4B-C). Rather, the degradation of Nsp9 by 9nb-SPOP_ΔNLS was inhibited to some extent by MG132 treatment (Fig. 4C). The significant inhibition of cytoplasmic Nsp9 degradation by CQ demonstrates that the degradation of Nsp9 by cytoplasmic distributed SPOP (SPOP_ΔNLS) involves pathways beyond the ubiquitin–proteasome system. This finding in consistent with a previous report that cellular stress induces the accumulation of SPOP in the cytoplasm and mediates the degradation of cytoplasmic substrates, but the mechanism has not yet been confirmed [25], which would be interesting for revealing the precise degradation pathway.

Fig. 4.

Autophagy–lysosomal pathway involved in 9nb-SPOP_ΔNLS−mediated degradation. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9 and 0, 1, or 3 µg of pcDNA3.1-Flag-9nb-SPOP_ΔNLS or control plasmid for 24 h, followed by treatment with 10 µM MG132 or 50 µM CQ for 6 h. (A) Western blot analysis and (B) flow cytometry analyses of the degradation efficiency of Nsp9 by 9nb-SPOP_ΔNLS in the presence of MG132 and CQ. The band intensities of (A) from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3), and Nsp9 was labeled with HA (Alexa Fluor™ 488) (B) and analyzed by a BD LSRFortessa X-20. (C) The fluorescence of (B) is presented as a column chart from three independent experiments. (D) 9nb-SPOP_ΔNLS-mediated degradation of Nsp9 regulates autolysosome formation. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9, pcDNA3.1-Flag-9nb-SPOP_ΔNLS and pmCherry-GFP-LC3B for 24 h, followed by treatment with 5 mM 3-methyladenine (3-MA) or 100 nM bafilomycin A1 (BAF A1) for an additional 6 h. The accumulation of LC3 is shown as red or green dots. The fluorescent signals were visualized using an LSM 980 Zeiss confocal microscope. Nuclei (blue) were labeled using DAPI. Scale bars, 10 μm. (E) 9nb-SPOP_ΔNLS facilitates the recruitment of p62 to Nsp9 in the cytoplasm. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9 and 3 µg of pcDNA3.1-Flag-9nb-SPOP, pcDNA3.1-Flag-9nb-SPOP_ΔNLS, pcDNA3.1-Flag-9nb-SPOP_mutΔNLS or control plasmid. After 24 h, the distribution of Nsp9 (green) and p62 (red) was detected by confocal microscopy. Nuclei (blue) were labeled using DAPI. Scale bars, 5 μm. Representative colocalization images are presented. Pearson’s correlation coefficient (PCC) was used to quantify the colocalization of Nsp9 with p62 using ImageJ software. All the experiments were performed three times, and the error bars indicate the SDs. Statistical significance was determined by two-way ANOVA; ****, P < 0.0001; ***, P < 0.001; *, P < 0.5; ns, P > 0.05

As CQ is a lysosomal autophagy inhibitor, the lysosome–autophagy proteolytic system is among the primary mechanisms maintaining protein homeostasis and the response to clear protein aggregates [63]. Autolysosome formation initiates the degradation of autophagosome cargo by exposing it to lysosomal hydrolases. Therefore, to visualize the autolysosomes and free autophagosomes, the autophagosome marker LC3 was fused to GFP and mCherry. The GFP signal is quenched when lysosomes fuse with autophagosomes because of the acid sensitivity of GFP [64]. We observed that in cells cotransfected with Nsp9 and 9nb-SPOP_ΔNLS, lysosomes fused with autophagosomes to form autolysosomes, which were detected on the basis of mCherry fluorescence (Fig. 4D). However, treatment with 5 mM 3-methyladenine (3-MA), an inhibitor of autophagosome formation, markedly reduced the accumulation of autolysosomes (Fig. 4D). Furthermore, the administration of bafilomycin A1 (BAF A1), which inhibits lysosomal acidification, led to the accumulation of LC3B-positive puncta with green fluorescence (Fig. 4D), indicating impaired autophagic flux. No increased accumulation of LC3 was detected in the 9nb-SPOP_mutΔNLS-cotransfected HEK293T cells with or without inhibitor treatment or in the Nsp9-transfected cells (Fig. S4). These results indicated that Nsp9 was degraded by 9nb-SPOP in the cytoplasm through the autophagy–lysosomal pathway.

To elucidate the molecular mechanism by which 9nb-SPOP_ΔNLS degraded Nsp9 through the autophagy–lysosomal pathway, we investigated the recruitment of the selective autophagy receptor p62 (SQSTM1). Upon binding to ubiquitinated substrates, it underwent conformational activation and oligomerization, subsequently translocating to the autophagosomal membrane to mediate the degradation of substrates. To determine whether 9nb-SPOP_ΔNLS-mediated Nsp9 ubiquitination recruited P62, the colocalization of Nsp9 and p62 was analyzed using confocal microscopy. p62 was robustly recruited to the Nsp9-9nb-SPOP_ΔNLS complex in the presence of functional 9nb-SPOP_ΔNLS, which strongly colocalized with Nsp9 (PCC = 0.96; Fig. 4E). In contrast, this recruitment was significantly reduced when the ubiquitination-defective mutant 9nb-SPOP_mutΔNLS was used, and cytoplasmic aggregates of Nsp9 and p62 were undetectable (PCC = 0.61; Fig. 4E). These results suggested that 9nb-SPOP_ΔNLS mediated Nsp9 cytoplasmic degradation through p62-mediated selective autophagy. Notably, in cells cotransfected with 9nb-SPOP, a small but distinct fraction of p62 formed cytoplasmic aggregates that colocalized with Nsp9 (PCC = 0.72; Fig. 4E). In striking contrast, despite the presence of Nsp9 aggregates in the nucleus, no recruitment of p62 was observed (Fig. S4C). This clear spatial distinction, which was consistent with our inhibitor assays (Fig. 3A), demonstrated that Nsp9 degradation was induced by 9nb-SPOP through both the proteasome and lysosome. These findings indicated that 9nb-SPOP mediated distinct degradation pathways in a subcellular compartment-specific manner.

Engineered 9nb-SPOP suppresses PRRSV replication in vitro

To demonstrate the anti-PRRSV activity of 9nb-SPOP, two Marc-145 cell lines stably expressing 9nb-SPOP and 9nb-SPOP_ΔNLS were established and named Marc-145-9nb-SPOP and Marc-145-9nb-SPOP_ΔNLS, respectively. Western blot analysis and fluorescence microscopy confirmed the successful expression of the bioPROTACs in stable cell lines, which were absent in wild-type Marc-145 controls (Fig. S5A, B), and the proliferation ability of the two stably expressing cell lines did not significantly differ from that of wild-type cells (Fig. S5C). As the antiviral activity was assessed using cell lines with constitutive bioPROTAC expression, conventional EC50 determination was not applicable. Next, the two stably expressing cell lines and wild-type Marc-145 cells were infected with the HuN4-EGFP recombinant virus at a multiplicity of infection (MOI) of 0.01. At 72 h postinfection (hpi), EGFP expression was notably reduced in both types of bioPROTAC-expressing cells, and only very small amounts of EGFP signal were detected in 9nb-SPOP_ΔNLS-expressing cells (Fig. 5A). In addition, the progeny virus titers revealed that the presence of 9nb-SPOP_ΔNLS significantly inhibited HuN4-EGFP replication at all the tested time points (P < 0.0001) (Fig. 5B).

Fig. 5.

Antiviral efficacy of bioPROTACs in vitro. (A) Marc-145, Marc-145-9nb-SPOP and Marc-145-9nb-SPOP_ΔNLS cell lines were infected with PRRSV HuN4-EGFP at an MOI of 0.01 for 72 h. The efficacy of the bioPROTACs against PRRSV infection was evaluated by IFA. PRRSV infection was indicated by EGFP expression. Scale bars, 1000 μm. (B) The supernatant in (A) was collected at 24, 48 and 72 h for virus titration in Marc-145 cells. (C–G) The three cell lines were infected with PRRSV HuN4-EGFP at MOIs of 0.1, 0.01, and 0.001 for 72 h, (C) the supernatant was collected for virus titration in Marc-145 cells, (D) the cells were lysed for RT–qPCR, the levels of PRRSV N were detected by Western blot analysis (E) and flow cytometry (F), and (G) the fluorescence of (F) is presented as a column chart from three independent experiments. (H–K) Evaluation of the antiviral efficacy of 9nb-SPOP_ΔNLS against multi-lineages of PRRSV. Marc-145 and Marc-145-9nb-SPOP_ΔNLS cell lines were infected with HP-PRRSV (HuN4), NADC30-like (strain HLJWK108), NADC34-like (strain LNTZJ1341) and genotype 1 (strain ZD1) at an MOI of 0.01 for 72 h, after which (H) the cell supernatants were used for virus titration, (I) the cells were lysed for RT–qPCR and (J) flow cytometry for PRRSV N expression detection. (K) The fluorescence of (J) is presented as a column chart from three independent experiments. (L) Analysis of 9nb-SPOP_ΔNLS-Fc entry into PAMs by IFA. PAMs were treated with 10 µg/mL 9nb-SPOP_ΔNLS-Fc for 2 h. The distribution of SPOP (red) was detected by confocal microscopy. Nuclei (blue) were labeled using DAPI. Scale bars, 10 μm. (M) Western blot analysis of 9nb-SPOP_ΔNLS-Fc entry efficiency in PAMs at different concentrations. PAMs were incubated with 9nb-SPOP_ΔNLS-Fc (0–50 µg/mL) for 2 h to evaluate dose-dependent cellular uptake. (N-P) Analysis of the anti-PRRSV activity of 9nb-SPOP_ΔNLS-Fc in vitro. PAMs were pretreated for 2 h with 9nb-SPOP_ΔNLS-Fc, Nnb-SPOP-Fc, a combination of 9nb-SPOP_ΔNLS-Fc and Nnb-SPOP-Fc (mixed at a 1:1 ratio) and NCnb-SPOP-Fc at a final concentration of 10 µg/mL. Afterward, the pretreated cells were infected with the PRRSV HuN4 strain at an MOI of 0.01 for 24 h. (N) The supernatant was collected for virus titration in Marc-145 cells, and the cells were used for (O) RT–qPCR and (P) IFA. PRRSV infection is indicated by N protein expression, and scale bars, 200 μm. All the experiments were performed three times, and the error bars indicate the SDs. Statistical significance was determined by one-way ANOVA or two-way ANOVA, ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.5; ns, P > 0.05

We further analyzed the effect of the viral infection dose on the antiviral effect of bioPROTACs. The two bioPROTAC-expressing cell lines were infected with the HuN4-EGFP recombinant virus at multiple infectious doses. After 72 hpi, the progeny virus titers, viral RNA synthesis levels and N protein expression levels were tested. The 9nb-SPOP_ΔNLS significantly decreased the virus titer by approximately 2 logs at different infectious doses, and the inhibitory effect of 9nb-SPOP_ΔNLS was markedly greater than that of 9nb-SPOP (P < 0.0001; Fig. 5C). Similar to the virus titer results, 9nb-SPOP_ΔNLS resulted in nearly 3 log and 2 log reductions in the number of intracellular PRRSV RNA copies at MOIs of 0.001 and 0.1, respectively (Fig. 5D). In addition, western blot analysis confirmed that the expression of N protein was suppressed by 9nb-SPOP_ΔNLS (Fig. 5E). Flow cytometry analysis revealed that 9nb-SPOP_ΔNLS reduced the number of PRRSV HuN4-infected cells, as indicated by the expression of N protein (Fig. 5F-G). The results confirmed that 9nb-SPOP_ΔNLS significantly reduced N protein expression in PRRSV HuN4-infected Marc-145 cells by approximately 100-fold at 0.1 MOI (P < 0.001) (Fig. 5G). These results demonstrated that the removal of NLSs could improve the ability of 9nb-SPOP to inhibit PRRSV proliferation.

Antiviral activity of 9nb-SPOP_ΔNLS against different PRRSV lineages

To assess the antiviral activity of 9nb-SPOP_ΔNLS, we evaluated its efficacy against three major PRRSV-2 lineages currently predominant in the epidemic in mainland China: highly pathogenic PRRSV (HP‑PRRSV, lineage L8E, HuN4 strain), NADC30‑like PRRSV (lineage L1C, HLJWK108 strain) [42], and NADC34‑like PRRSV (lineage L1A, LNTZJ1341 strain) [65]. These three lineages account for more than 95% of currently circulating PRRSV‑2 in China. In addition, we further analyzed the antiviral activity of the bioPROTACs against PRRSV‑1 (ZD1 strain) [66]. Infection of stable 9nb-SPOP_ΔNLS-expressing Marc-145 cells with the indicated PRRSV strains (MOI = 0.01) revealed an antiviral effect against multi-lineages of PRRSV. Viral titer analysis at 72 hpi demonstrated a significant inhibitory effect, with virus titer log reductions (95% CI) of 2.5 (2.1–2.8), 1.6 (1.3–2.0) and 1.7 (1.3–2.1) for the HuN4, HLJWK108 and ZD-1 strains, respectively (P < 0.0001; Fig. 5H). In addition, 9nb-SPOP_ΔNLS significantly reduced the number of genomic copies of all the PRRSV genotypes tested (P < 0.0001; Fig. 5I). Notably, the inhibitory effect on viral transcription was more pronounced for the HuN4 and HLJWK108 strains than for the other strains. Flow cytometric analysis revealed that 9nb-SPOP_ΔNLS significantly reduced the proportion of PRRSV-infected cells. In addition, it more effectively inhibited the HuN4 and HLJWK108 strains than the LNTZJ1341 and ZD1 strains (P < 0.0001; Fig. 5K). These results demonstrated that 9nb-SPOP_ΔNLS exerted inhibitory effects on different PRRSV species and lineages.

BioPROTAC-Fc significantly inhibits PRRSV replication in PAMs

PAMs with Fcγ receptors served as the primary cells for PRRSV infection in pigs. To improve the efficiency of bioPROTAC entry into PAMs, porcine IgG Fc fragments were fused with bioPROTACs as a delivery tag (Fig. S6A). The expression and purification of 9nb-SPOP_ΔNLS, Nnb-SPOP, Nsp9nb, and Nnb fused to the porcine Fc domain were analyzed by SDS–PAGE and Western blot analysis (Fig. S6B–D). To assess the cellular uptake of bioPROTAC-Fc fusion proteins, PAMs were treated with 9nb-SPOP_ΔNLS-Fc or 9nb-SPOP_ΔNLS and analyzed via IFA. IFA revealed that 9nb-SPOP_ΔNLS-Fc was widely distributed in PAMs following treatment with 10 µg/mL for 2 h (Fig. 5L). Furthermore, western blot analysis demonstrated that 9nb-SPOP_ΔNLS-Fc was internalized by PAMs in a dose-dependent manner after 2 h of incubation (Fig. 5M).

Next, the effects of bioPROTAC-Fc fusion proteins on the inhibition of PRRSV replication in PAMs were analyzed. The PAMs were treated with 9nb-SPOP_ΔNLS-Fc, Nnb-SPOP-Fc, a combination of 9nb-SPOP_ΔNLS-Fc and Nnb-SPOP-Fc (mixed at a 1:1 ratio) and NCnb-SPOP-Fc at a final concentration of 10 µg/mL for 2 h. Then, the bioPROTAC-Fc pre-treated cells were infected with the PRRSV HuN4 strain at 0.01 MOI for 24 h. RT–qPCR and virus titer assays consistently demonstrated that all Fc-fusion bioPROTACs, except NCnb-SPOP-Fc, significantly reduced PRRSV RNA levels and infectious viral titers (P < 0.0001). Notably, the combined administration of 9nb-SPOP_ΔNLS-Fc and Nnb-SPOP-Fc resulted in a significant reduction in viral titers by 2.4 logs (95% CI: 1.8–3.1) and in viral RNA copy numbers by 3.6 logs (95% CI: 3.2–4.0) relative to those of the PRRSV infection control. This combination exhibited a significantly greater inhibitory effect compared with either 9nb-SPOP_ΔNLS-Fc or Nnb-SPOP-Fc treatment alone (P < 0.0001) (Fig. 5N and O). Furthermore, confocal microscopy revealed that the combined administration of 9nb-SPOP_ΔNLS-Fc and Nnb-SPOP-Fc resulted in the lowest level of N protein expression, which was markedly reduced compared with that in all the other groups (Fig. 5P). This pronounced reduction suggested that a synergistic effect might be obtained by simultaneously interfering with multiple stages of the PRRSV replication cycle.

BioPROTACs exhibit a high genetic barrier to antiviral resistance

To evaluate the ability of bioPROTACs in overcoming acquired resistance caused by target mutations, PRRSV was serially passaged in the Marc-145-9nb-SPOP_ΔNLS cell line. Control passages were conducted using parental Marc-145 cells in the absence or presence of 0.2 mM ribavirin [52]. Viruses from passages P1, P5, P10, and P15 were collected for viral titration and viral genome copy number. Both 9nb-SPOP_ΔNLS and 0.2 mM ribavirin significantly suppressed PRRSV replication from the first passage (Fig. 6A–D, P < 0.0001). In the presence of 9nb-SPOP_ΔNLS, the viral titer and genome copy number of HuN4-bioPROTAC progressively decreased across passages. Both the viral titer and viral genome copy number of P15 were reduced by 4.214 log10 and 4.468 log10, respectively, compared to HuN4 P0 (P < 0.0001, Fig. 6B and D). Notably, pressure selection with ribavirin resulted in undetectable infectious viral particles from P5 to P10, and the viral genome copies were at an extremely low level (Fig. 6C). However, viral replication rebounded after P10, and the progeny viral production at P15 reached a level comparable to that of P0 (P > 0.05, Fig. 6B).

Fig. 6.

Assessment of the genetic resistance barrier of bioPROTAC (9nb-SPOP_ΔNLS) in vitro. The PRRSV HuN4 strain was consecutively passaged to P15 in the Marc-145-9nb-SPOP_ΔNLS cell line or in Marc-145 cells treated with or without 0.2 mM ribavirin (RBV). (A) Viral titers and (C) viral genome copy numbers were determined at indicated passages. (B–D) The heatmap of multiple changes and statistical differences of (B) viral titers and (D) viral genome copy numbers following serial passage. The heatmap represents the mean of log10 fold changes relative to HuN4 P0. The * represents statistical differences relative to HuN4 P0 (**, P < 0.01 and ****, P < 0.0001). (E) Mutations detected in the Nsp9 of PRRSV HuN4 P15 virus under selection with bioPROTAC (9nb‑SPOP_ΔNLS) or RBV. The numbers in parentheses indicate the nucleotide or amino acid position in Nsp9. (F) Sequence alignment of PRRSV HuN4 at P10 and P15 with or without selection pressure. The dots represent identical amino acids with the reference sequences. The alignment was conducted by the Clustal W method in MEGALIGN software (DNASTAR Lasergene). (G) Co-IP analysis for the binding of mutant Nsp9FT to bioPROTAC (9nb-SPOP_ΔNLS). HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9FT and or control plasmids (pCAGGS-HA) with pcDNA3.1-Flag-9nb-SPOP_ΔNLS. After 24 h, the cells were subjected to IP with an antibody against Flag, followed by Western blot analysis. (H) Western blot analysis of bioPROTAC (9nb-SPOP_ΔNLS) mediated degradation of mutant Nsp9FT. HEK293T cells were cotransfected with 1 µg of pCAGGS-HA-Nsp9FT and 3 µg of pcDNA3.1-Flag-9nb-SPOP_ΔNLS or control plasmids (pcDNA3.1-Flag) for 24 h. Band intensities from three independent experiments were quantified using ImageJ software and are presented as the mean ± SD (n = 3). The Marc-145 cells and Marc-145-9nb-SPOP_ΔNLS cells were infected with HuN4-RVB P15 at MOI of 0.1 for 72 h. (I) The supernatant was collected for viral titration in Marc-145 cells, and (J) the cells were lysed for RT–qPCR. Statistical significance was determined using the unpaired two-tailed Student’s t test, ****, P < 0.0001; **, P < 0.01

Next, we assess the genetic stability of PRRSV Nsp9 under distinct selection pressures by isolating the viral RNA from the P10 and P15 of the HuN4-bioPROTAC, HuN4-RBV, and HuN4 WT. The Nsp9 gene was amplified and subsequently sequenced for mutation analysis. The consensus sequences were derived from 3 independent clones. Sequence alignment revealed that all amino acid changes detected in HuN4-bioPROTAC-P15 and HuN4-WT-P15 were silent mutations. In contrast, HuN4-RBV-P15 acquired an Ala to Thr mutation at position 612 (A612T) located within the predicted Nsp9 RdRp thumb domain. An additional Leu to Phe (L6F) mutation was identified in the N-terminal region of Nsp9, which is a domain thought to overlap with that of Nsp8 (Fig. 6E and F) [67]. Notably, neither of these mutations was present in the P10 viruses (Fig. 6F). This was consistent with the observation that the replication of HuN4-RBV resistant variants significantly recovers after P10.

To assess whether the ribavirin-selected mutations L6F and A612T could result in resistance to bioPROTAC-mediated degradation, we constructed an Nsp9 mutant plasmid (pCAGGS-HA-Nsp9FT) harboring these two amino acid residues. Co-IP confirmed that 9nb-SPOP_ΔNLS maintained binding to mutant Nsp9FT (Fig. 6G). Following co-transfection in HEK293T cells, 9nb-SPOP_ΔNLS induced pronounced degradation of Nsp9FT (Fig. 6H), with an efficiency comparable to that observed in wild-type Nsp9 degradation. Furthermore, in vitro antiviral assays confirmed that 9nb‑SPOP_ΔNLS expression in Marc‑145 cells significantly suppressed the replication of the ribavirin‑resistant variant HuN4‑RBV‑P15 (Fig. 6I and J, P < 0.0001). These results suggested that bioPROTAC-based strategies possess a high genetic barrier to the emergence of antiviral resistance.

BioPROTACs mediate degradation of target proteins in mice

Next, we studied the in vivo degradation efficacy of 9nb-SPOP_ΔNLS and Nnb-SPOP in mice, and the TransIT-EE (enhanced) hydrodynamic delivery protocol was employed for plasmid-mediated gene transfer in mice, which was analyzed by an in vivo imaging system [68]. A schematic diagram of target protein degradation in mice is shown in Fig. 7A. Prior to in vivo imaging, purified Nsp9-mCherry and N-mCherry proteins were validated, and both mCherry-fused proteins showed high-intensity fluorescence, with no signal detected in the blank tube (NC) or delivery solution (Fig. 7B). Whole-body fluorescence imaging at 24 h postinjection revealed a significant reduction in Nsp9-mCherry and N-mCherry signal intensity in mice coexpressing 9nb-SPOP_ΔNLS and Nnb-SPOP degraders, respectively (Fig. 7C-c and C-i). The degradation efficacy is shown in Fig. 7D and E, with 9nb-SPOP_ΔNLS and Nnb-SPOP leading to a greater than 50% reduction in the total fluorescence intensity of Nsp9-mCherry and N-mCherry compared with that of the controls (P < 0.01). In contrast, coexpression with the nontargeting NCnb-SPOP control did not affect the Nsp9-mCherry or N-mCherry signal (P > 0.05). All control groups showed only background-level signals (Figs. 7C-a, e, f, g, k, l). Notably, the degradation effect was more pronounced in lung tissues, and compared with mCherry-fused protein controls, both 9nb-SPOP_ΔNLS and Nnb-SPOP caused a greater than 80% reduction in fluorescence intensity (Fig. 7F–H). Similarly, the levels of the target proteins Nsp9 and N significantly decreased in the lungs upon 9nb-SPOP_ΔNLS and Nnb-SPOP treatment (Fig. 7I-J). These findings confirmed that both 9nb-SPOP_ΔNLS and Nnb-SPOP induce efficient in vivo degradation of target proteins, with pronounced efficacy in the lungs. These conclusions lay the foundation for further antiviral applications of bioPROTACs targeting key PRRSV viral proteins in vivo.

Fig. 7.

In vivo target degradation efficacy of bioPROTACs. (A) Schematic illustration of the degradation experiment conducted in mice. (B) To establish optimal fluorescence imaging conditions, two purified prokaryotically expressed mCherry-tagged target proteins (Nsp9-mCherry/N-mCherry) were placed in EP tubes, with blank tubes and delivery solution serving as negative controls. (C-J) The distribution of mCherry fluorescence in the mice was assessed. Plasmid mixtures were prepared by combining 10 µg of either the pCAGGS-HA-Nsp9-mCherry or pCAGGS-HA-N-mCherry plasmid with 20 µg of pcDNA3.1-Flag-9nb-SPOP_ΔNLS or pcDNA3.1-Flag-Nnb-SPOP, respectively. The mixtures were subsequently administered to the mice via tail vein injection. mCherry fluorescence was detected 24 h postinjection, and (C) the total fluorescence intensity was quantified using IndiGO 2 software. Statistical analysis was performed on the total fluorescence intensity of Nsp9-mCherry (D) and N-mCherry (E) in mice. (F) The lungs were collected and imaged ex vivo. Statistical analysis was performed on the total fluorescence intensity of Nsp9-mcherry (G) and N-mcherry (H) in the lungs of mice. Western blot analysis of the degradation levels of Nsp9 (I) and N (J) proteins in the lungs of mice. All the fluorescence images (B, C, F) were obtained by a small animal in vivo imaging system (Berthold NightOWL LB 983, Germany) at appropriate wavelengths (λex = 500 nm, λem = 600 nm) and scaled to the same minimum and maximum values. Bar graphs (D, E, G, H) show the mean ± SD values of the samples, which were calculated via Student’s t test; ***, P < 0.001; **, P < 0.01; ns, P > 0.05

Delivery of bioPROTACs via mRNA-LNPs suppresses PRRSV replication and transmission in vivo

To overcome the delivery limitations of bioPROTACs, we administered bioPROTACs to piglets in two forms: bioPROTACs-mRNA-LNP and bioPROTACs-Fc. Both strategies facilitated cell-membrane penetration, thereby enabling an evaluation of the therapeutic potential in a PRRSV-infected piglet model. On the basis of the above results, we used combined Nsp9nb-SPOP_ΔNLS mRNA and Nnb-SPOP mRNA (1:1) to generate mRNA-LNP (mRNA(N + 9)), and mixed purified Nsp9nb-SPOP_ΔNLS-Fc and Nnb-SPOP-Fc (1:1) (Protein(N + 9)) to perform animal experiments; Protein(NCnb) served as a negative control. Piglets infected with the HP-PRRSV HuN4 strain were treated with drugs every two days for a total of three administrations (Fig. 8A). Daily rectal temperature, mortality, and lung pathology were monitored throughout the experimental period. Clinical scores were assessed at 0 and 7 dpi. All surviving pigs were euthanized at 10 dpi for organ collection and subsequent analysis. At 1 and 4 dpi, the mean rectal temperatures in the Mock, Protein(NCnb), and Protein(N + 9) groups exceeded 40.5 °C, indicating a biphasic fever pattern. These groups maintained a persistent high fever from 4 dpi. Piglets in the Protein(NCnb) and Mock groups died between 5 and 7 dpi. In contrast, the mRNA(N + 9) group displayed a shortened pyrexic phase, which lasted only two days, with body temperature returning to normal by 7 dpi (Fig. 8B). The NC group remained normal throughout the experiment. A significant survival benefit was observed in the mRNA(N + 9) group compared with the controls (Fig. 8C). All the piglets in the Mock and Protein(NCnb) groups died by 7 dpi, with a median survival of 6 days. In contrast, the mRNA(N + 9) treatment resulted in a significant survival benefit, with a 75% survival rate (3/4). This improvement was statistically confirmed by the log-rank test (χ²=6.943, df = 1, P = 0.0084). The median survival time of the Protein(N + 9) group was 7.5 days, and one patient survived (25%) at the study endpoint. However, this difference did not reach statistical significance compared with the Mock group (χ²=3.543; df = 1; P = 0.0585). All the piglets in the NC group survived. Clinical scores at 7 dpi were significantly greater than those at 0 dpi in the Mock and Protein(NCnb) groups (scores > 15, P < 0.0001). The Protein(N + 9) group also showed significantly aggravated clinical signs at 7 dpi (P = 0.0054). The clinical signs in the mRNA(N + 9) group were notably attenuated, however, the mean score remained significantly higher than that at 0 dpi (P = 0.0159) (Fig. 8D).

Fig. 8.

Antiviral efficacy of bioPROTACs in piglets. (A) Schematic illustration of the animal experimental protocol. (B) Rectal temperature was recorded daily for surviving animals in all groups. (C) Survival curves of piglets in each group are shown, and P values were determined using the log-rank (Mantel-Cox) test. (D) Clinical scores for all the animals were compiled on days 0 and 7. (E) Representative lung images. Samples from all groups were collected at death or 10 dpi. (F) The gross pathological changes found in the piglets of each group were quantified using a scoring (100-point) system. (G) Microscopic lung lesions. Lung tissue samples were fixed, embedded, and stained with H&E to facilitate the observation of pathological changes. Representative images were captured, and the scale bars represent lengths of 50 and 100 μm. The images are representative of three independent biological replicates. (H) Viral loads in representative tissues of all experimental animals. (I) Viral load in blood and shedding in nasal and anal swabs. Samples were collected at 1 and 4 dpi. Viral copy numbers for each pig are presented. Bar graphs (D, H) and scatter plots (F, I) show the mean ± SD values of the samples, with P values as indicated, which were calculated via one-way ANOVA (F) and two-way ANOVA with Dunnett’s multiple-comparison test (D, H, I). ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; ns, P > 0.05

Throughout the study, gross lung lesions were evaluated in both dead and surviving pigs using a standardized lesion scoring system to quantitatively assess pathological alterations. Severe pulmonary consolidation and edema were observed in the Mock and Protein(Nbn) groups (Fig. 8E), with no significant difference in lesion scores between these two groups (P = 0.6229) (Fig. 8F). In the Protein(N + 9) group, three piglets died at 7 and 8 dpi. The lung of the survivors also exhibited pulmonary consolidation and hemorrhage, though these lesions were less severe than those in the Mock group (P = 0.0517) (Fig. 8E-F). In contrast, compared with the Mock controls, the pigs in the mRNA(N + 9) group showed only mild diffuse hemorrhage with no other notable lung pathologies (Fig. 8E). Furthermore, the lung lesion scores in both the mRNA(N + 9) and NC groups (P = 0.0073 and P = 0.0002, respectively) were significantly lower than those in the Mock group (Fig. 8F), suggesting that treatment with mRNA (N + 9) effectively alleviated the pathological changes associated with PRRSV infection. Histopathological examination revealed severe interstitial pneumonia and lobular consolidation following PRRSV infection in the Mock group. These pathological features were markedly alleviated in pigs treated with Protein(N + 9) or mRNA(N + 9). Moreover, the Protein(NCnb) group showed no improvement in inflammatory cell infiltration or alveolar septal thickening (Fig. 8G). Additionally, viral loads in major organs were quantified by RT–qPCR to evaluate the protective efficacy of bioPROTACs. PRRSV was widely detected in tissues such as lung, liver, heart, spleen, and kidney, as well as in multiple lymph nodes and immune organs. Treatment with mRNA(N + 9) significantly suppressed PRRSV replication across all the examined organs (P < 0.0001; Fig. 8H). Notably, Protein(N + 9) exhibited the most pronounced reduction in viral load in the lung (P < 0.0001), while the reductions observed in the kidney and inguinal lymph nodes did not reach statistical significance (ns, Fig. 8H). These results demonstrated that both mRNA(N + 9) effectively suppressed PRRSV replication in pigs. To assess the impact of bioPROTACs on early viral dissemination, we quantified viremia and viral shedding by testing blood, nasal, and anal swabs collected at 1 and 4 dpi. RT–qPCR analysis revealed high levels of PRRSV replication at 4 dpi. Compared with the Mock group, the mRNA( N+ 9) treatment group presented significantly lower PRRSV loads in blood and anal shedding at 1 dpi (P < 0.01) and a pronounced reduction in viral loads across all three organs at 4 dpi (P < 0.0035) (Fig. 8I). The Protein(N + 9) treatment significantly suppressed viremia and anal shedding at both time points (P < 0.05) but did not significantly affect nasal shedding (Fig. 8I). These results suggested that both mRNA(N + 9) and Protein(N + 9) could reduce early viremia and mucosal shedding, thereby potentially limiting the transmission of PRRSV during the critical initial phase of infection.

Discussion

PROTACs offer a promising strategy to develop antivirals for RNA viruses with error-prone replication and high mutation rates [69], presenting an alternative to traditional small-molecule drugs or antibody drugs. This degradation-based approach can achieve therapeutic outcomes that go beyond those of conventional inhibition. The small molecule-derived PROTAC-mediated degradation Nef protein has not only suppressed HIV-1 replication but also restored T-cell immunity by upregulating CD4 and MHC-I expression [70]. Nevertheless, the development of effective PROTACs remains challenging for viruses that currently lack small-molecule inhibitors, thereby limiting the development of novel antivirals. In this study, we developed modular nanobody-based bioPROTACs and used clinically significant PRRSV as a viral model to systematically evaluate the antiviral effects of bioPROTACs against PRRSV. Importantly, these modular bioPROTACs enabled flexible retargeting through nanobody substitution, showing antiviral activity against different PRRSV lineages and significantly inhibiting PRRSV replication in vivo (Fig. 9).

Fig. 9.

Nanobody-based bioPROTACs targeting the degradation of key PRRSV proteins suppress PRRSV replication and have activity against different PRRSV lineages. (1) A modular nanobody-based bioPROTAC platform in which a nanobody is conjugated to the SPOP BTB domain enabled rapid antiviral drug development through the substitution of target-specific nanobodies. (2) LNP-based nanoparticles were utilized to encapsulate and deliver bioPROTAC mRNA for in vivo expression. (3) BioPROTACs induce autophagolysosomal degradation of PRRSV RdRp in the cytoplasm and (4) ubiquitin–proteasome degradation of the PRRSV N protein in the nucleus

The nonstructural proteins of PRRSV form a highly coordinated functional network during viral replication. The core catalytic components of this network are encoded by ORF1b and include the Nsp9, Nsp10 and Nsp12, which together constitute the functional core of the viral replication and transcription complex (RTC) [71, 72]. The correct assembly and localization of these catalytic proteins rely on a membrane scaffold formed by the transmembrane proteins Nsp2, Nsp3, and Nsp5 [73]. These scaffold proteins remodel the endomembrane system of the host cell and specifically recruit nonstructural proteins such as Nsp9 to membrane-associated RTC sites [74]. Although Nsp9 has been reported to localize to the nucleus with a diffuse pattern in the cytoplasm, during viral infection, Nsp9 is specifically recruited to membrane-bound RTC in the cytoplasm to execute its replicase function [71]. To achieve targeted degradation of Nsp9, this study used SPOP, a typical nuclear-localized E3 ubiquitin ligase adaptor, as an E3 recruiting component. Our results revealed that although Nsp9 exhibits nuclear localization and both 9nb-SPOP and its nuclear localization signal deficient variant 9nb-SPOP_ΔNLS degraded Nsp9 with comparable efficiency, compared with 9nb-SPOP, 9nb-SPOP_ΔNLS significantly suppressed PRRSV replication in two different genotypes. This finding clearly indicates that degradation of cytoplasmic Nsp9 is more critical for effectively inhibiting PRRSV replication, which may offer a novel strategy for PRRSV control.

PRRSV Nsp9 has an RdRp structure, which significantly affects the viral replication efficiency and its fatal virulence in piglets [27, 71]. Although amino acid changes may occur, its genes are highly conserved compared with those of other nonstructural proteins [75, 76]. BioPROTAC technology offers a higher genetic barrier to resistance compared with traditional small-molecule antiviral drugs. Unlike ribavirin, which exerts low selective pressure for resistance leading to the emergence of resistant variants in RNA viruses such as PRRSV [53], continuous pressure selection with 9nb-SPOP_ΔNLS did not induce any amino acid mutations in the Nsp9 protein. To evade bioPROTAC-mediated degradation, a viral variant would need to acquire multiple, specific mutations that must concomitantly abrogate nanobody binding, preserve protein functionality, and prevent instability-driven degradation. The probability of such a combination of mutations arising concurrently is low. Our study suggests that such bioPROTAC degraders can also be applied to other viruses dependent on RdRp activity, such as severe acute respiratory syndrome coronavirus (SARS-CoV) [77] and equine arteritis virus (EAV) [78].

SPOP is widely recognized as a canonical nucleus-localized adapter for Cullin3 E3 ubiquitin ligase. Our study revealed that the subcellular localization of both SPOP and viral proteins influenced the degradation pathway. To target the degradation of the cytoplasmic Nsp9 protein, we engineered a SPOP variant (9nb-SPOP_ΔNLS). This construct mediated the polyubiquitination of Nsp9 and facilitated its degradation primarily through the autophagy-mediated lysosomal pathway. We found that 9nb-SPOP_ΔNLS induced the formation of dynamic, liquid-like compartments containing Nsp9 in the cytoplasm, which is essential for the generation of aggregates of ubiquitinated substrates and facilitates the recognition of ubiquitinated Nsp9 by the autophagy receptor p62, leading to autophagosome formation [79]. This mechanism is similar to AUTAC-mediated targeted degradation of impaired mitochondria, which also relies on K63-linked ubiquitination of the target protein [80]. Notably, removing the NLS from SPOP prevented excessive nuclear accumulation of the target-bioPROTAC complex, thereby increasing the cytoplasmic availability of the degrader and its antiviral efficacy against PRRSV. These findings indicated that although MATH-BTB proteins such as SPOP primarily functioned within the ubiquitin–proteasome system, cellular relocalization could redirect their activity toward alternative degradation pathways, such as autophagy. This plasticity in degradation pathway engagement, which depends on cellular compartmentalization, represents a novel aspect of PROTAC design.

An ideal antiviral drug should have high virus specificity and avoid off-target effects. However, the development of small-molecule PROTACs against viruses requires an extensive screening process because of the limited availability of small-molecule ligands of the viral protein and the few known efficient E3 ligases that cooperate with small-molecule ligands [81]. In contrast, replacing the substrate recognition region of the E3 ligase with an antibody or peptide allows the rapid design of bioPROTAC degraders, especially for viruses with no known small-molecule inhibitors [11]. We have established a modular SPOP-derived bioPROTAC platform for antiviral development. This approach is similar to emerging technologies such as LYTAC and AUTAC [82, 83] but offers unique advantages through its nanobody-mediated targeting specificity and subcellular localization control. Its efficacy can be rapidly redirected against new targets by substituting the nanobody, while controlling alteration of the degrader’s localization allows targeted protein degradation in either the cytoplasm or the nucleus. This flexibility enables the rational design and systematic screening of highly specific antiviral drugs against rapidly mutating viruses.

Neutralizing antibodies can effectively block viral entry; however, mutations in neutralizing epitopes on highly variable viral structural proteins can lead to neutralization “escape” [84]. In contrast, conserved antigenic regions are ideal antiviral targets, avoiding immune escape caused by antigenic variation [29, 85, 86]. These antibodies may not have direct antiviral function but usually have high affinity. In this study, the nanobodies we employed are non-neutralizing antibodies, but the Nsp9 nanobody shows modest antiviral activity, whereas the N protein nanobody shows no antiviral activity [46]. We found that compared with nanobody alone, nanobody-SPOP significantly enhanced anti-PRRSV activity, although direct antiviral function was lacking. Nanobodies serve as target-binding ligands in bioPROTAC, which does not require their intrinsic antiviral activity, and their ability to mediate efficient protein degradation relies on high‑affinity and specific binding to the target protein. This strategy demonstrates the capacity of the bioPROTAC platform to convert non-antiviral antibodies into antiviral agents. Therefore, this strategy not only expands the application scope of non-neutralizing antibodies but also reduces the risk of viral escape caused by antigenic variation.

Currently, bioPROTACs pose challenges for clinical translation because of their large molecular size, poor solubility, and limited bioavailability [34, 35]. To ensure optimal antiviral efficacy in vivo, piglets were pre-treated with the bio-macromolecular drug one day prior to viral challenge. However, the Fc-fusion protein still demonstrated limited protective effects (Fig. 8), which stands in contrast to its antiviral activity observed in vitro (Figs. 5N-P). The disparity between in vitro and in vivo outcomes is potentially due to a combination of factors, including poor tissue targeting specificity, inadequate tissue penetration of the fusion protein, and renal clearance [87], all of which could limit therapeutic drug concentrations at the lungs, the primary site of PRRSV replication. However, when delivered via mRNA-LNP, bioPROTACs markedly suppressed viral replication and shedding, which improved survival in piglets, indicating that the therapeutic effect of bioPROTACs relied on drug delivery systems. The animal experimental protocol employed in this study also has certain limitations. Although prophylactic administration ensures that the drug reaches effective concentrations in the early stages of infection, this administration does not represent the therapeutic antiviral treatment commonly encountered in clinical practice. Future study should include postinfection dosing regimens to enable a more accurate assessment of its clinical translational potential.

Beyond the mRNA-LNP used in this study, recent advances in PROTAC delivery include the use of polymeric nanoparticles, inorganic nanoparticles, and protein-based nanoparticles, which have shown promise in preclinical cancer models and could be applied for antiviral development [88, 89, 90, 91, 92, 93]. An engineered bioPROTAC template that complexes with cationic and ionizable lipids for cytosolic delivery enables bioPROTAC access to various subcellular compartments [93]. Furthermore, surface functionalization of these nanocarriers enhances their ability for targeted delivery, which improves therapeutic efficacy while reducing systemic toxicity [94, 95]. However, despite these functional advantages of nanocarriers, the synthetic complexity also poses significant challenges. Thus, the combination of nanotechnology and PROTACs constitutes a promising strategy for accelerating the clinical application of bioPROTAC drugs [96].

Conclusion

In summary, this study successfully established a modular nanobody-based bioPROTAC platform that enabled rapid antiviral drug construction through the substitution of target protein-specific nanobodies. The platform efficiently degraded key PRRSV RdRp and N proteins both in vitro and in vivo while demonstrating antiviral activity against multi-lineages of PRRSV. When delivered via the mRNA-LNP system, the combined application of bioPROTACs significantly decreased the piglet mortality rate to 25%. Importantly, the subcellular localization of both the target protein and bioPROTACs determined the degradation pathway, confirming that the cytoplasm-localized 9nb-SPOP_ΔNLS degraded Nsp9 through the autophagy–lysosome pathway. This study expands the applicability of SPOP-derived bioPROTACs from nuclear proteins to cytoplasmic proteins, providing a novel strategy for the development of antiviral therapies against highly variable viruses.

Author contributions

*S.S., M.S. and H.W. contributed equally to this work. F.M., T.A. and Y.L. designed the experiment. S.S., X.S., H.Z., S.C. and M.S. performed the experiments. S.S., H.W., X.S. and S.W. analyzed the data. W.Y., N.Z. and Y.Y. prepared the figures. M.S., Y.T., K.Z. and Z.T. provided experimental material. S.S., J.C., X.C., Y.L. and F.M. wrote the manuscript. All the authors revised the manuscript.

Funding

This study was supported by grants from the National Key Research & Development Program of China (No. 2022YFD1800300 to TA), Basic Research Center, Innovation Program of Chinese Academy of Agricultural Sciences (CAAS-BRC-LPDC-2025-02 to FM), the Foundation of the National Research Center of Engineering and Technology for Veterinary Biologicals (No. GTKF(23)008 to FM), the China Postdoctoral Science Foundation (No. 2025M773034 to SS), the Natural Science Foundation of Heilongjiang Province (No. ZD2023C005 to TA), the Innovation Program of the Chinese Academy of Agricultural Sciences (CAAS-CSLPDCP-202301 to TA), and the Heilongjiang Provincial Natural Science Foundation of China (No. LH2023C023 to WY).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments were performed within animal biosafety level 2 (ABSL-2) facilities under oversight and were approved by the Animal Ethics Committee of the Harbin Veterinary Research Institute. The animal ethics committee approval numbers were 231207-03-GR for the mouse experiments and 250220-06-GR for the piglet experiments. Consent to Participate declaration: not applicable.

Consent for publication

Consent to Publish declaration: not applicable.

Competing interests

Tongqing An and Fandan Meng filed a patent related to this technology.