ABSTRACT

   As an essential component of host defense against infection,
   necroptosis is a novel highly regulated mode of cell death that is
   mediated by signaling complexes containing receptor-interacting protein
   kinase 1 (RIPK1) and RIPK3. Lymphocyte depletion and immunosuppression
   are typical clinical features of pigs infected with classical swine
   fever virus (CSFV). Here, we provide the first evidence for the
   involvement of necroptosis in the necrosis of T lymphocytes in the
   spleen and peripheral blood of pigs infected in vivo with CSFV.
   However, it is true for some viruses with non-cytopathic effects,
   including CSFV, to balance host defense against infection. In vitro,
   the induction of autophagy by CSFV at a later stage of infection
   clearly restricts necroptosis. Mechanistic studies revealed that CSFV
   NS4A protein promoted tripartite motif-containing 25 expression,
   synergistically induced the occurrence of mitophagy, targeted the
   autophagic degradation of RIPK3 to block the progression of necroptosis
   occurrence, and achieved persistent viral infection. Interestingly, we
   found that RIPK3 was able to specifically localize at the outer
   mitochondrial membrane, and the autophagy receptor NDP52 was most
   likely involved in the autophagic degradation of RIPK3 during CSFV
   infection. Our findings provide evidence supporting that the
   CSFV-induced autophagy pathway plays an important role in counteracting
   host cell necrosis, enriching our knowledge of pathogens that may
   subvert and evade necroptosis this host defense, and shedding new light
   on understanding the mechanisms of T lymphocyte exhaustion and
   immunosuppression during CSFV infection.

IMPORTANCE

   CSFV infection in pigs causes persistent high fever, hemorrhagic
   necrotizing multi-organ inflammation, and high mortality, which
   seriously threatens the global swine industry. Cell death is an
   essential immune response of the host against pathogen invasion, and
   lymphopenia is the most typical clinical feature in the acute phase of
   CSFV infection, which affects the initial host antiviral immunity. As
   an “old” virus, CSFV has evolved mechanisms to evade host immune
   response after a long genetic evolution. Here, we show that necroptosis
   is a limiting host factor for CSFV infection and that CSFV-induced
   autophagy can subvert this host defense mechanism to promote its
   sustained replication. Our findings reveal a complex link between
   necroptosis and autophagy in the process of cell death, provide
   evidence supporting the important role for CSFV in counteracting host
   cell necrosis, and enrich our knowledge of pathogens that may subvert
   and evade this host defense.

   KEYWORDS: autophagy, necroptosis, CSFV, NS4A, TRIM25, mitophagy,
   autophagic degradation, RIPK3

INTRODUCTION

   Necroptosis (also known as cell-programmed necroptosis) emerges as a
   novel signaling molecule-regulated, orderly manner of cell death that
   retains some features of apoptosis, such as being genetically
   controlled, and exhibits necrosis-like morphological features, such as
   inflammatory responses that can cause surrounding tissues ([61]1,
   [62]2). In the absence of apoptotic conditions (mainly when caspase 8
   is inhibited), necroptosis can be triggered by multiple signaling
   pathways, including the tumor necrosis factor receptor family (TNFR),
   the Toll-like family of receptors, and the cytoplasmic DNA sensor
   Z-DNA-binding protein 1 (ZBP1/DAI), among others ([63]3 [64]– [65]5).
   Receptor-interacting protein kinase 1 (RIPK1) and RIPK3 form a protein
   complex called “necrosome” through their respective RIP homotypic
   interaction motif (RHIM) domains, activated RIPK3 recruits and
   activates mixed lineage kinase domain-like pseudokinase (MLKL) ([66]6),
   and the oligomerized MLKL translocates from the cytoplasm to the cell
   membrane and organelle membranes to form a permeability pore, causing
   cell necrosis ([67]7, [68]8).

   As a component of host defense against infection, necroptosis is
   closely related to the body’s innate immune response, the development
   and progression of inflammation-related diseases, and so on ([69]9
   [70]– [71]11). This contribution was first derived from the study of
   vaccinia virus (VV), in which mice were infected with VV and
   specifically induced the formation of an RIPK1-RIPK3 necrosome complex
   in the liver, triggering tissue necroptosis and inflammation, leading
   to elevated viral replication and mouse mortality ([72]10).
   Subsequently, it has been documented that upon infection of host cells
   with influenza A virus (IAV), ZBP1 recognizes the IAV genome and in
   turn activates host RIPK3 kinase, which induces cell death through the
   MLKL-driven necroptosis pathway ([73]12, [74]13). In addition, bovine
   parvovirus, murine cytomegalovirus, and reovirus, among others, can
   activate RIPK3-dependent necroptosis ([75]14, [76]15). The spread of
   the virus depends on the state of the infected cell, as well as the
   interactions between the virus and eukaryotic cells. Cells die after
   infection with the virus and prevent virus propagation, whereas some
   viruses encode suppressor genes of cell death to counteract this
   protective mechanism ([77]16 [78]– [79]18). Herpes simplex virus type 1
   (HSV1) depends on the cognate RHIM suppressor to prevent ZBP1
   activation by newly synthesized RNA in human cells ([80]19). In
   addition, Epstein-Barr virus-encoded LMP1 promotes K63-linked
   polyubiquitination of RIPK1 and inhibits protein expression, while
   inhibiting K63-linked polyubiquitination of RIPK3 prevents necroptosis
   in human nasopharyngeal epithelial and nasopharyngeal carcinoma cells
   ([81]20). Necroptosis is important for the innate immune control of
   virus-induced inflammation and viral infection, while correspondingly,
   viruses have also evolved counterpart strategies to evade the host
   immune response ([82]21).

   It is generally accepted that active cell death modalities mainly
   include apoptosis, necroptosis, and autophagy. As a highly conserved
   intracellular metabolic process in eukaryotes, autophagy delivers
   damaged organelles, misfolded or aggregated proteins, and invading
   pathogens to lysosomes for degradation or recycling ([83]22). Selective
   trafficking of autophagy substrates requires multiple cargo receptor
   mediation, including sequestosome-1 (SQSTM1/P62), nuclear dot protein
   52 (CALCOCO2/NDP52), optineurin (OPTN), neighbor of BRCA1 gene 1
   (NBR1), Tax1-binding protein 1 (TAX1BP1), Toll-interacting protein
   (TOLLIP), etc. ([84]23 [85]– [86]25). Autophagy receptors, which
   recognize a variety of substrates (damaged organelles such as
   mitochondria, ubiquitinated or non-ubiquitinated proteins) and directly
   bind microtubule-associated protein 1 light chain 3 (MAP1LC3/LC3) for
   lysosomal degradation, are hubs linking autophagy substrates to
   lysosomes ([87]22, [88]26). Both autophagy and necroptosis are highly
   regulated and both balance cell death and survival, and although
   several reports have suggested reciprocal crosstalk and overlap between
   them ([89]27 [90]– [91]29), to date the mechanistic details remain
   poorly understood.

   Classical swine fever virus (CSFV), which is the same Flaviviridae as
   hepatitis C and dengue viruses, is an enveloped RNA virus and the
   etiological agent of swine fever (CSF) ([92]30). The mature CSFV
   particle structure includes four structural proteins [core protein (C)
   and envelope glycoprotein (E^rns, E1, and E2)] and eight non-structural
   proteins (N^pro, P7, NS2, NS3, NS4A, NS4B, NS5A, and NS5B) ([93]31
   [94]– [95]37). CSFV infection of pigs causes persistent high fever,
   hemorrhagic necrotizing multi-organ inflammation with high mortality,
   and seriously threatens the global swine industry. As an “old” virus,
   CSFV has undergone a long genetic evolution and has evolved related
   mechanisms to escape the host immune response ([96]38, [97]39). All
   leukocyte populations in peripheral blood, especially lymphopenia, are
   the most typical clinical features in the acute phase of CSFV
   infection, thereby affecting the initial host antiviral immunity.
   Furthermore, interestingly, when CSFV was propagated in host cells in
   vitro, there was no obvious cytopathic effect. Although many studies
   related to CSFV replication have been performed, the mechanism of viral
   escape from this disease is unknown.

   Previously, we have found that CSFV can induce autophagy and utilize
   autophagic vesicles for replication ([98]40, [99]41), and furthermore,
   CSFV was able to induce mitochondrial fission and mitophagy to promote
   its infection process ([100]42). In this study, we found that CSFV
   infection in vivo was able to activate necroptosis in peripheral blood
   mononuclear cells (PBMCs) and spleen, leading to lymphocyte necrosis.
   Whereas, in contrast to this behavior, CSFV clearly inhibited
   necroptosis at a later stage of infection in vitro. Further mechanistic
   studies revealed that the CSFV NS4A protein interacts with the E3
   ubiquitin ligase tripartite-motif containing 25 (TRIM25) ([101]43,
   [102]44), promoted its expression, and targeted it to mediate
   autophagic degradation of RIPK3 and thereby blocked the progression of
   necroptosis genesis, achieving persistent viral infection.
   Interestingly, we found that RIPK3 was able to specifically localize at
   the outer mitochondrial membrane, and the autophagy receptor NDP52 was
   most likely involved in the autophagic degradation of RIPK3 during CSFV
   infection. Our results provide evidence supporting an important role
   for CSFV NS4A in counteracting host cell necrosis, enrich our knowledge
   of pathogens that may subvert and evade this host defense, and shed new
   light on understanding the mechanisms of T lymphocyte exhaustion and
   immunosuppression during CSFV infection.

RESULTS

Peripheral blood T lymphopenia caused by CSFV infection of piglets was
associated with necroptosis

   CSFV has a strong tropism for epithelial cells and immune cells and is
   a typical immunosuppressive virus that seriously damages the
   hematopoietic and immune systems ([103]45). First, we investigated the
   effect of CSFV infection on the number of lymphocytes and their
   subsets, and we found that the proportion of lymphocytes among
   peripheral blood mononuclear cells in infected piglets transiently rose
   to 73.33% at 1 dpi from 65.44% before the challenge and decreased to
   6.00% at 6 dpi, where CD3^+ and CD4^+ T lymphocytes decreased to 2.53%
   and 1.33% from 19.86% and 4.15% before the challenge, respectively
   ([104]Fig. 1A and B). Experimental piglets were sacrificed at 7 dpi for
   pathological examination and specimen collection. We further observed
   the histopathological changes in the spleen, and the red and white pulp
   of the spleen were well demarcated in uninfected control pigs, the
   splenic corpuscles were structurally intact and regularly shaped, and
   the central artery was clearly visible (black arrow).

Fig 1.

   [105]Fig 1
   [106]Open in a new tab

   CSFV-infected piglets cause peripheral blood T lymphocyte reduction and
   spleen necrosis. (A) Peripheral blood mononuclear cells were isolated
   from uninfected and CSFV Shimen strain-infected piglets at days 1, 4,
   and 6 to label the proportions of CD3^+ and CD4^+ T lymphocytes,
   respectively, by flow cytometry in the representative images. (B)
   Quantification of the ratio of CD3^+and CD4^+ T lymphocytes shown in
   panel A. (C) The experimental pigs were sacrificed at 7 dpi, and
   spleens were collected for tissue preparation and stained with
   hematoxylin and eosin, and then spleen’s histopathological changes were
   analyzed with a NIKON Eclipse Ci biological microscope (Japan), under
   the magnification of 200× and 400×. Bar: 100 and 200 µm, respectively.
   Histopathological changes as indicated by differently colored arrows
   and dotted circles can be seen in the panel.

   Swelling of splenic trabeculae and more rounded cavities (black arrows)
   in the cytoplasm of smooth muscle cells were observed in the
   CSFV-infected group of pigs. Splenic corpuscles were sparse and small
   in size, the boundaries of most splenic corpuscles were already
   indistinct, and splenic marginal zone and red pulp hemorrhages were
   severe. Lymphocytes are greatly reduced in number and lightly
   pigmented. The wall of the splenic artery showed hyalinosis, with
   thickening and homogeneous red staining of the wall and compaction of
   the lumen (yellow arrow). Focal disintegration of lymphocytic necrosis
   with infiltration of granulocytes around necrotic cells is seen in the
   white pulp (black dotted circle). In addition, the splenic sinusoids
   were markedly dilated and visible with an excess of packed red blood
   cells (red arrows) ([107]Fig. 1C). Consistent with previous reports,
   CSFV infection of spleens induced the development of severe
   hemorrhagic, necroinflammation ([108]46). These data suggest that CSFV
   infection caused extensive lymphocyte depletion and tissue necrosis;
   however, the underlying mechanism was unclear and controversial.

   Lymphopenia is a major cause of immunosuppression. To explore the cause
   of lymphopenia in piglets with CSFV infection, we performed
   transcriptomic analysis of CD3^+ T lymphocytes in the peripheral blood
   as well as splenic lymphocytes (SPLs) of piglets infected with CSFV,
   respectively, and set up uninfected CD3^+ T lymphocytes as well as SPL
   controls. The DEGs were determined by the results of |log[2] fold
   change| >1 and P [adj] <0.05 by RNA sequencing. Compared with the
   control group, 293 genes were upregulated, and 161 genes were
   downregulated in the CD3^+ T lymphocyte-treated group ([109]Fig. 2A);
   831 genes were upregulated, and 962 genes were downregulated in the
   SPL-treated group ([110]Fig. 2B). Significant differential genes
   related to cell death, including MLKL, an executive protein of
   programmed necrosis, were selected among the DEGs in both cell types;
   IL1A, a pleiotropic cytokine that accompanies the onset of
   inflammation; CASP1, a caspase involved in the execution phase of
   apoptosis. Cluster analysis showed that the CSFV-infected group had a
   good cluster-like relationship with the control group, indicating that
   the CSFV infection-induced gene transcriptome differences within CD3^+
   T lymphocytes and spleen lymphocytes were small but significantly
   different between groups ([111]Fig. 2C and D). The associated genes
   were subjected to pathway association analysis, which revealed a
   cellular network linking CSFV infection to lymphocyte death ([112]Fig.
   2E). Go functional annotation of DEGs showed that the most
   significantly enriched biological processes were defense response (Go:
   0006952), and 66 DEGs were enriched. The most significant enrichment in
   molecular function was for chemokine receptor binding (Go: 0042379),
   with enrichment of 10 DEGs. The most significantly enriched gene in the
   cellular component was plasma membrane part (Go: 0044459), and 59 DEGs
   were enriched ([113]Fig. 2F). The results of Kyoto Encyclopedia of
   Genes and Genomes (KEGG) analysis showed significant enrichment of
   cytokine-cytokine receptor interaction, Toll-like receptor signaling
   pathway, IL-17 signaling pathway, TNF signaling pathway, viral protein
   interaction with cytokines and cytokine receptors, and necroptosis
   pathway ([114]Fig. 2G). We further selected a large number of enriched
   cell death-related genes to validate the accuracy of the transcriptome
   data, respectively ([115]Fig. 2H and I), and the results showed that
   the quantitative reverse transcription polymerase chain reaction
   (qRT-PCR) was approximately the same as the RNA-seq results, indicating
   the reliability of the transcriptome data. All these data suggest that
   these pathways may play important roles in porcine peripheral blood
   CD3^+ T lymphocytes against CSFV infection and that CSFV
   infection-induced lymphopenia may be closely related to the
   significantly enriched necroptosis pathway.

Fig 2.

   [116]Fig 2
   [117]Open in a new tab

   Global omics analysis of CSFV-infected peripheral blood CD3^+ T
   lymphocytes and splenic lymphocytes. (A and B) Volcano plots of
   CSFV-infected peripheral blood CD3^+ T lymphocytes and spleen
   lymphocytes, respectively, to visually display the overall distribution
   of differential genes. (The DEGs with P-value <0.05, |log2FC| >1 were
   selected by R package software. Scatterplots represent individual
   genes; gray, genes with no significant difference; yellow,
   significantly differentially upregulated genes; purple, significantly
   differentially downregulated genes.) (C and D) Differential gene
   clustering plots of CSFV-infected peripheral blood CD3^+ T lymphocytes
   and spleen lymphocytes, respectively. According to the similarity of
   gene expression situation in each sample, the genes were clustered to
   visually demonstrate the difference of gene expression situation in
   different samples. (Each column represents a sample, and each row
   represents a gene. Red, upregulated genes; blue, downregulated genes.)
   (E) CSFV infects cellular networks associated with lymphocyte death.
   (Octagons represent pathways, circles represent genes, and network
   visualization generated by Cytoscape v.3.8.0.) (F) The Go functional
   annotation of DEGs in CSFV-infected peripheral blood CD3^+ T
   lymphocytes (green, biological process; blue, molecular function; red,
   cellular component). (G) The KEGG pathway enrichment analysis of DEGs
   from CSFV-infected peripheral blood CD3^+ T lymphocytes was performed
   based on KEGG pathways as background, the significance level of gene
   enrichment of each pathway was calculated by Fisher’s exact test, and
   the top 30 enriched pathways were selected to draw a bubble plot. (H)
   PBMCs from CSFV-infected piglets at 7 dpi were isolated; RNA was
   extracted, and significantly enriched differential gene mRNA fold
   changes were quantified by qRT-PCR, and bars were plotted together with
   fold change from RNA-seq. (I) Spleens of CSFV-infected piglets at 7 dpi
   were collected; RNA was extracted, and significantly enriched
   differential gene mRNA fold changes were quantified by qRT-PCR, and
   bars were plotted together with fold change from RNA-seq.

CSFV infection induces necroptosis in the spleen and PBMCs in vivo

   To explore whether CSFV infection could cause necrosis in peripheral
   blood and spleen lymphocytes, by transmission electron microscopy, we
   detected an increased number of cells with distinct necrotic features,
   indicated by their enlarged cell volume, focal disruption of the cell
   membrane, plasma membrane leakage, and cell lysis ([118]Fig. 3A and C),
   as well as a significantly increased number of autophagosomes
   ([119]Fig. 3A and B), in PBMCs isolated from CSFV-infected piglets.
   Furthermore, excessive erythrocyte flooding (red arrows) was observed
   in the spleens of CSFV-infected piglets, consistent with ([120]Fig.
   1C). Similarly, both necrotic cells and autophagosome numbers were
   obviously increased ([121]Fig. 3D through F), suggesting that there may
   be some association between CSFV infection-induced autophagy and
   necrotic cells in vivo. We further examined the protein levels of
   necroptosis markers, first determining the successful infection of
   PBMCs by CSFV through the expression of CSFV N^pro protein and found
   that MLKL expression level, an executive protein of necroptosis, was
   significantly higher in PBMC cells from CSFV-infected piglets, along
   with inflammatory cytokines TNF-α and IL-1β were significantly
   expressed ([122]Fig. 3G), indicating that the induction of inflammation
   by CSFV infection is likely accompanied by the onset of necroptosis.
   Furthermore, we found that CSFV infection significantly increased the
   mRNA levels of necroptosis hallmark genes RIPK1, RIPK3, and MLKL in
   PBMCs ([123]Fig. 3I). In addition, mRNA levels of TNF-α, TNFR1, and
   ZBP1, upstream genes of necroptosis, were significantly increased in
   CSFV-infected PBMCs, whereas little effect on Caspase-8 mRNA levels was
   observed ([124]Fig. 3K). Meanwhile, CSFV infection was found to be able
   to promote the proportion of necrotic cells in PBMCs by flow cytometry
   analysis ([125]Fig. 3H). Similarly, we detected significantly higher
   mRNA levels of MLKL and ZBP1 in the spleen of CSFV-infected piglets,
   whereas no significant changes in mRNA levels of RIPK1, RIPK3, and
   TNF-α were observed ([126]Fig. 3J and L). This may be attributed to
   significant differences among the RNA samples within the group.
   Furthermore, the high expression of necroptosis marker proteins RIPK1,
   RIPK3, and MLKL in the spleen of CSFV-infected piglets was validated by
   immunohistochemistry (IHC; [127]Fig. 3M through O). The above results
   indicated that CSFV infection was able to induce PBMCs and spleen to
   undergo necroptosis in vivo. To explore whether CSFV-induced
   necroptosis has a tissue tropism, we examined the mRNA levels of
   necroptosis marker genes in different immune tissues, such as tonsil,
   lymph node, and thymus, and non-immune tissues, such as lung and
   kidney, respectively, and found that the expression of these genes was
   significantly promoted by CSFV in the thymus, significantly suppressed
   in kidney and tonsil, and unchanged in lung and lymph node (Fig. S1).
   The results suggest that the occurrence of necroptosis may be related
   to the viral load and the extent of the lesions in the infected tissue,
   with some tissue tropism, and also reflect to some extent that
   necroptosis as a viral defense mechanism may play an important role in
   combating CSFV infection.

Fig 3.

   [128]Fig 3
   [129]Open in a new tab

   Necroptosis is induced in PBMCs and spleens of CSFV-infected piglets in
   vivo. (A) PBMCs from CSFV-infected and uninfected piglets at 7 dpi were
   isolated, and necrotic cell features as well as autophagy-like vesicles
   were visualized by transmission electron microscopy (N, nucleus; M,
   mitochondria; Go, golgi; ASS, autophagosome). (B) Quantitative analysis
   of the number of autophagosomes observed in (A). Error bars indicate
   the mean (± SD) of five independent experiments. **, P < 0.01 (one-way
   ANOVA). (C) Quantitative analysis of the proportion of necrotic cells
   observed in (A). Error bars indicate the mean (± SD) of three
   independent experiments. ***, P < 0.001 (one-way ANOVA). (D) Spleens
   from CSFV-infected and uninfected piglets at 7 d were collected, and
   necrotic cell features as well as autophagy-like vesicles were
   visualized by transmission electron microscopy (N, nucleus; red arrow,
   red blood cell). (E) Quantitative analysis of the number of
   autophagosomes observed in (D). Error bars indicate the mean (± SD) of
   five independent experiments. *, P < 0.05 (one-way ANOVA). (F)
   Quantitative analysis of the proportion of necrotic cells observed in
   (D). Error bars indicate the mean (± SD) of three independent
   experiments. *, P < 0.05 (one-way ANOVA). (G) The protein levels of
   RIPK1, MLKL, TNF -α, IL-1β, and CSFV N^pro in PBMC of CSFV-infected and
   non-infected piglets were detected by Western blot. (I and K) The mRNA
   expressions of RIPK1, RIPK3, MLKL, TNF-α, TNF-R1, ZBP1, and Caspase-8
   in PBMC of CSFV-infected and non-infected piglets were determined by
   qRT-PCR. The fold changes were related to the internal control β-actin.
   Error bars indicate the mean (± SD) of three independent experiments.
   **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 (two-way ANOVA). (J and
   L) The mRNA expressions of RIPK1, RIPK3, MLKL, TNF-α, and ZBP1 in
   spleens of CSFV-infected and non-infected piglets were determined by
   qRT-PCR. The fold changes were related to the internal control β-actin.
   Error bars indicate the mean (± SD) of three independent experiments.
   ***, P < 0.001 and ****, P < 0.0001 (two-way ANOVA). (H) Flow cytometry
   was used to determine the proportion of necrotic cells in PBMCs from
   CSFV-infected and uninfected piglets at 7 d. (M) The expression of CSFV
   E2 in the spleen of infected and uninfected piglets was detected by
   IHC. (N) The expression of RIPK1, RIPK3, and MLKL in the spleen of
   infected and uninfected piglets was detected by IHC. During this
   period, we fixed, embedded, and sectioned spleen tissues. The sections
   were then stained with specific antibodies against RIPK1, RIPK3, and
   MLKL, respectively. The appearance of similar contours in the sliced
   images is a result of the proximity of the tissue sections. (O)
   Immunohistochemical scoring of the detection indices in (N).

Necroptosis is induced at the early stage of CSFV infection but inhibited at
the late stage in vitro

   We found that CSFV infection-induced necroptosis has a certain tissue
   tropism, so what is the relationship between CSFV infection and
   necroptosis in vitro? Is this consistent in vivo? We cultured the
   isolated PBMCs in vitro and monitored the protein level and mRNA level
   changes of necroptosis marker genes RIPK1, RIPK3, MLKL, and ZBP1 at 24,
   48, and 60 h after CSFV infection; meanwhile, we determined the
   successful establishment of a PBMC model of CSFV infection in vitro by
   detecting the expression of CSFV N^pro protein. Compared with control
   cells not infected with CSFV, CSFV infection for 24 h promoted the
   expression of necroptosis-related proteins, while it markedly inhibited
   the expression of RIPK1-, RIPK3-, MLKL-, and ZBP1-related proteins in
   PBMCs at 48 or 60 h ([130]Fig. 4A and B). With longer infection time,
   CSFV could obviously increase RIPK3 and ZBP1 mRNA levels in PBMCs
   compared with the control, whereas it suppressed MLKL mRNA levels at 60
   h of infection ([131]Fig. 4G). This observation may appear
   contradictory to the results at the protein level but is consistent
   with the conclusion that CSFV infection can significantly induce
   necrotic apoptosis in PBMCs. As for the protein-level inhibition
   observed in the later stages, it is likely closely related to the
   autophagy process induced by CSFV. Furthermore, using an in vitro
   established cell model of CSFV infection of PK-15 and 3D4/21, similar
   to infected PBMC cells, we found that CSFV was able to promote the
   expression of necroptosis-related genes in PK-15 ([132]Fig. 4C, D, and
   H) and 3D4/21 ([133]Fig. 4E, F, and I) cells at the preinfection stage
   and inhibit the expression of related genes at the later stage of
   infection as the infection time progressed. These data indicate that
   CSFV was able to induce necroptosis in the early phase of in vitro
   infection, whereas it was inhibited in the middle and late phases. The
   differential changes in necrotic apoptosis-related gene and protein
   levels induced by CSFV infection in PBMCs, PK-15, and 3D4/21 cells may
   be closely associated with the tissue tropism of CSFV. This may also be
   an important reason why CSFV infection in vitro fails to elicit
   cytopathic effects.

Fig 4.

   [134]Fig 4
   [135]Open in a new tab

   Necroptosis was induced at the early stage and inhibited at the late
   stage of CSFV infection in vitro. To establish a PBMC cell model of
   CSFV infection in vitro, cells were collected at 24, 48, and 60 h
   post-infection, and protein expression and mRNA levels of necroptosis
   markers including RIPK1, RIPK3, MLKL, and ZBP1 were detected by Western
   blot (A) and qRT-PCR (G). (B) Gray-scale value analysis of the
   detection indexes in (A); the relative intensity was related to the
   internal control Tubulin. Error bars indicate the mean (± SD) of three
   independent experiments. ns, P ≥ 0.05*; P < 0.05; **, P < 0.01; and
   ***, P < 0.001 (two-way ANOVA). To establish a PK-15 cell model of CSFV
   infection in vitro, cells were collected at 24, 48, and 60 h
   post-infection, and protein expression and mRNA levels of necroptosis
   marker genes including RIPK1, RIPK3, MLKL, and ZBP1 were detected by
   Western blot (C) and qRT-PCR (H). (D) Gray-scale value analysis of the
   detection indexes in (C); the relative intensity was related to the
   internal control Tubulin. Error bars indicate the mean (± SD) of three
   independent experiments. ns, P ≥ 0.05*; P < 0.05; **, P < 0.01; and
   ***, P < 0.001 (two-way ANOVA). To establish a 3D4/21 cell model of
   CSFV infection in vitro, cells were collected at 24, 48, and 60 h
   post-infection, and protein expression and mRNA levels of necroptosis
   markers including RIPK1, RIPK3, MLKL, and ZBP1 were detected by Western
   blot (E) and qRT-PCR (I). (F) Gray-scale value analysis of the
   detection indexes in (E); the relative intensity was related to the
   internal control Tubulin. Error bars indicate the mean (± SD) of three
   independent experiments. ns, P ≥ 0.05*; P < 0.05; **, P < 0.01; and
   ***, P < 0.001 (two-way ANOVA).

CSFV NS4A interacts with RIPK3 and TRIM25 to inhibit necroptosis

   In order to explore the molecular mechanism by which necroptosis is
   inhibited by CSFV infection in vitro, we overexpressed the major core
   protein (C) as well as the non-structural proteins P7, NS4A, NS4B, and
   NS5A of CSFV in vitro and observed significant colocalization of only
   GFP-NS4A and HA-RIPK3 in HEK-293T cells co-transfected with HA-RIPK3
   and pGFP-C1, GFP-P7, GFP-C, GFP-NS4A, GFP-NS4B, GFP-NS5A, respectively
   ([136]Fig. 5A), and RIPK interacted with CSFV NS4A protein was further
   verified by co-immunoprecipitation (Co-IP) assay; additionally, the
   protein level of RIPK3 was obviously decreased in the presence of NS4A
   ([137]Fig. 5C). Meanwhile, we transfected PK-15 cells with plasmids
   expressing different viral proteins and found that RIPK3 protein levels
   were decreased, and MLKL protein was also decreased in the presence of
   CSFV NS4A alone compared with transfection with empty vector pEGFP-C1
   ([138]Fig. 5B). Previous studies in our laboratory identified TRIM25,
   an E3 ubiquitin ligase, as a potential interacting molecule with CSFV
   NS4A protein by mass spectrometry, and we also further demonstrated the
   relevant functional relationship by Co-IP ([139]Fig. 5D and 2A) and
   laser confocal assay (Fig. S2B). Notably, overexpression of TRIM25
   promoted the protein expression of NS4A ([140]Fig. 5D). Moreover, NS4A
   specifically mediated the decrease in RIPK3, which was more pronounced
   when NS4A levels were gradually increased (Fig. S2C). Interestingly, as
   NS4A levels increased, so did TRIM25 expression (Fig. S2C). It was
   previously reported that TRIM25 was able to interact with RIPK3 and
   mediate its degradation, acting as a brake molecule for necroptosis
   ([141]44), then is this phenomenon also present in PK-15 cells? We
   found that RIPK3 was gradually degraded as the expression level of
   TRIM25 increased (Fig. S2D), and there was a significant spatial
   colocalization of CSFV NS4A with TRIM25, RIPK3 ([142]Fig. 5E).
   Furthermore, significant co-localization of NS4A with TRIM25 (Fig. S3A)
   and NS4A with RIPK3 (Fig. S3B) was observed during CSFV infection.
   Taking a step further, we transfected the CSFV NS4A-expressing plasmid
   into PK-15 cells overexpressing TRIM25 and knocking down TRIM25,
   respectively, to explore whether NS4A specifically mediates the
   degradation of RIPK3 via TRIM25. The results showed that the
   overexpression of TRIM25 specifically promoted the degradation of RIPK3
   by NS4A, whereas RIPK1 levels were not affected by TRIM25 ([143]Fig.
   5F). Correspondingly, the knockdown of TRIM25 was able to specifically
   restore the inhibitory effect of NS4A on RIPK3, and neither NS4A nor
   TRIM25 had an obvious regulatory effect on RIPK1 (Fig. S2E). Moreover,
   it was not difficult to find that CSFV NS4A modulated the expression
   level of the necroptosis executioner MLKL via TRIM25 ([144]Fig. 5F and
   2E).

Fig 5.

   [145]Fig 5
   [146]Open in a new tab

   CSFV NS4A interacts with RIPK3 and TRIM25 to inhibit necroptosis. (A)
   HEK-293T cells were co-transfected with HA-RIPK3 together with pGFP-C1,
   GFP-P7, GFP-C, GFP-NS4A, GFP-NS4B, or GFP-NS5A for 24 h, fixed cells
   were incubated with anti-HA tag primary antibody and then stained with
   Alexa fluor 555-conjugated anti-Rabbit-IgG secondary antibody (red),
   and nuclei were stained with DAPI. Scale bars: 25 µm. (B) pGFP-C1,
   GFP-P7, GFP-C, GFP-NS4A, GFP-NS4B, and GFP-NS5A were transfected into
   PK-15 cells for 24 h, and the cells were lysed. The expression levels
   of viral proteins, RIPK1, RIPK3, and MLKL, were detected by Western
   blot. The right panel was gray-scale value analysis, the relative
   intensity was related to the internal control GAPDH. Error bars
   indicate the mean (± SD) of three independent experiments. ****, P <
   0.0001 (two-way ANOVA). (C) FLAG-RIPK3 was co-transfected with pEGFP-C1
   and GFP-NS4A, respectively, into PK-15 cells, and the cells were lysed;
   Co-IP assay and Western blot analysis were performed. (D) Myc-TRIM25
   was co-transfected with pEGFP-C1 and GFP-NS4A in HEK-293T cells, while
   a control group of cells was set up in which pCMV-N1-Myc was
   co-transfected with pEGFP-C1 and GFP-NS4A, cells were lysed; Co-IP
   experiments and Western blot analysis were performed. (E) PK-15 cells
   were co-transfected with three constructs, HA-RIPK3, Myc-TRIM25 with
   GFP-NS4A, and a control construct of cells co-transfected with
   HA-RIPK3, Myc-TRIM25 with pEGFP-C1, HA-RIPK3, pCMV-N1-Myc with
   GFP-NS4A, Myc-TRIM25, pCAGGs-HA with GFP-NS4A, The fixed cells were
   incubated with anti-HA primary antibody and anti-Myc primary antibody,
   respectively, and then stained with Alexa fluor 555-conjugated
   anti-Rabbit-IgG Mouse antibody (red) and Dylight 405-labeled
   anti-Rabbit-IgG secondary antibody (blue) (left panel). Scale bars: 10
   µm. The fluorescence intensity profiles of HA-RIPK3 (blue),
   GFP-NS4A(green), and Myc-TRIM25(red) were analyzed using Image J plugin
   and plotted by GraphPad Prism 9 (right panel). (F) Myc-TRIM25 was
   co-transfected with pEGFP-C1 and GFP-NS4A in PK-15 cells, while a
   control group of cells was set up in which pCMV-N1-Myc was
   co-transfected with pEGFP-C1 and GFP-NS4A; the cell lysates were
   detected by Western blot. The gray-scale value analysis was related to
   the internal control GAPDH (lower panel). Error bars indicate the mean
   (± SD) of three independent experiments. ****, P < 0.0001 (two-way
   ANOVA).

CSFV infection inhibits necroptosis via autophagy/mitophagy in vitro

   Previously, we found that CSFV was able to induce necroptosis at the
   initial stage of infection, which was inhibited at the later stage.
   Previous studies have found that after CSFV infects host cells, the
   replication of viruses can be promoted by activating the
   autophagy/mitophagy pathway in cells, and the utilization of autophagic
   vesicles provides a favorable site for the replication of viruses
   ([147]40, [148]42). Necroptosis, as an effective defense mechanism of
   the organism, is activated at the initial stage of CSFV infection. So,
   during the interaction between CSFV and the host organism, is it
   possible for CSFV to interfere with or even inhibit necroptosis through
   the autophagy pathway, thus achieving persistent infection within the
   cells?

   PK-15 ([149]Fig. 6A) cells uninfected and infected with CSFV for 48 h
   were, respectively, treated with autophagy agonist rapamycin. The
   results showed that rapamycin elevated the autophagic flux in PK-15
   cells and promoted the expression of CSFV N^pro protein. More
   importantly, rapamycin further strengthened the inhibitory effect of
   CSFV infection on necroptosis-related proteins, especially RIPK3 and
   MLKL. On this basis, we further measured the expression levels of
   mitophagy-related proteins and key molecules of necroptosis by treating
   CSFV-uninfected and -infected PK-15 ([150]Fig. 6B) cells with mitophagy
   inducer carbonyl cyanide 3-chlorophenylhydrazone (CCCP). The results
   showed that CCCP increased the mitophagy level in PK-15 cells and
   further strengthened the inhibitory effect of CSFV infection on
   necroptosis key molecules, especially RIPK3 and MLKL. In addition, we
   further verified this phenomenon using flow cytometry by pretreating
   CSFV-infected PK-15 cells with necroptosis inducer (TSZ; TNF-α, Smac
   mimetic, and z-VAD-Fmk) autophagy agonist rapamycin ([151]Fig. 6C).
   Furthermore, we verified that RIPK3 expression was significantly
   elevated in the presence of CSFV infection and inhibited the autophagy
   pathway in CSFV-infected cells by introducing inhibitors at different
   stages of autophagy or transfecting with shRNA targeting interfering
   autophagy-related genes ([152]Fig. 6D). These results suggest that CSFV
   most likely inhibits necroptosis by activating the autophagy pathway.

Fig 6.

   [153]Fig 6
   [154]Open in a new tab

   CSFV inhibits necroptosis through autophagy/mitophagy and
   necroptosis-restricted CSFV infection in vitro. (A) PK-15 cells were
   infected with CSFV (MOI = 1) for 48 h, rapamycin was treated for 6 h,
   and an uninfected cell control was set up. Cell lysates were then
   detected by Western blot (left panel). Relative intensity related to
   internal control Tubulin was analyzed using Image J plugin and plotted
   by GraphPad Prism 9 (right panel). Error bars indicate the mean (± SD)
   of three independent experiments. ns, P ≥ 0.05; *, P < 0. 1; **, P <
   0.01; ***, P < 0.001; ****, P < 0.0001 (two-way ANOVA). (B) PK-15 cells
   were infected with CSFV (MOI = 1) for 48 h, CCCP treated for 6 h, and
   an uninfected cell control was set up. Cell lysates were then detected
   by Western blot (left panel). Relative intensity related to internal
   control Tubulin was analyzed using Image J plugin and plotted by
   GraphPad Prism 9 (right panel). Error bars indicate the mean (± SD) of
   three independent experiments. ns, P ≥ 0.05; *, P < 0. 1; ***, P <
   0.001; ****, P < 0.0001 (two-way ANOVA). (C) PK-15 cells were infected
   with CSFV (MOI = 1) for 48 h and treated with TSZ (20 ng/mL TNF-α, 100
   nM Smac mimetic, and 20 µM z-VAD-Fmk) alone or TSZ together with
   rapamycin for 6 h, and a control of uninfected cells was set up.
   Necrotic cells were identified by flow cytometry with PI and Annexin V
   staining. A representative plot of the data is shown in the left panel,
   and the mean (± SD) of three independent experiments is shown in the
   right panel. *, P < 0. 1 and ****, P < 0.0001 (two-way ANOVA). (D)
   PK-15 cells infected with CSFV (MOI = 1) were transfected with
   Flag-RIPK3 for 24 h and then treated with 3-methyladenine (3-MA, 5 mM),
   chloroquine (CQ, 20 µM), or bafilomycin A1 (Baf A1, 100 nM) for 6 h,
   and cell lysates were detected by Western blot (upper panel).
   Flag-RIPK3 was co-transfected with shNC, shATG5, shBECN1, and shLC3
   into CSFV-infected PK-15 cells, and cell lysates were detected by
   Western blot (lower panel). (E) PK-15 (left panel) and 3D4/21 (right
   panel) cells infected with CSFV (MOI = 1) were transfected with
   HA-RIPK3 or MLKL for 48 h, and pCAGGs-HA was transfected with a control
   group. The cell supernatants were collected for TCID50 assay. Error
   bars indicate the mean (± SD) of three independent experiments. *, P <
   0.1 and **, P < 0.01 (one-way ANOVA). (F) PK-15 cells infected with
   CSFV (MOI = 1) were treated with TSZ for 6 h, and control of uninfected
   cells was set up. The mRNA expressions of RIPK1, RIPK3, MLKL, and CSFV
   NS5B in the cell lysates were determined by qRT-PCR. The fold changes
   were related to the internal control β-actin. Error bars indicate the
   mean (± SD) of three independent experiments. *, P < 0.1; **, P < 0.01;
   ***, P < 0.001 (one-way ANOVA). (G) PK-15 cells infected with CSFV (MOI
   = 1) were treated with increasing amounts of TSZ (wedge shaped, 0; 20
   ng/mL TNF-α, 100 nM Smac mimetic, and 10 µM z-VAD-Fmk; 40 ng/mL TNF-α,
   100 nM Smac mimetic, and 10 µM z-VAD-Fmk; 40 ng/mL TNF-α, 100 nM Smac
   mimetic, and 20 µM z-VAD-Fmk) for 6 h, and the cell lysates were
   detected by Western blot. (H) PK-15 cells were treated with increasing
   amounts of necrosulfonamide (NSA; wedge shaped, 0, 1, 5, and 10 µM),
   and the cell lysates were detected by Western blot (upper panel). PK-15
   cells infected with CSFV (MOI = 1) were treated with NSA (5 µM), and
   dimethyl sulfoxide (DMSO) was treated with a control group. The cell
   lysates were detected by Western blot (lower panel). (I) PK-15 (left
   panel) and 3D4/21 (right panel) cells were infected with CSFV (MOI = 1)
   for 48 h and treated with TSZ (20 ng/mL TNF-α, 100 nM Smac mimetic, and
   20 µM z-VAD -Fmk) alone or TSZ together with rapamycin for 6 h, and a
   control of uninfected cells was set up. The cell lysates were detected
   by Western blot.

Necroptosis restricted CSFV infection, while induction of autophagy by CSFV
antagonizes necroptosis for persistent infection

   Exploring the role of necroptosis in CSFV infection, we found that the
   reduction in CSFV infection as determined by released infectious virus
   titers was observed in PK-15 and 3D4/21 cells overexpressing RIPK3 and
   MLKL ([155]Fig. 6E). Correspondingly, we treated CSFV-infected PK-15
   cells with TSZ, the relative expression of CSFV NS5B markedly decreased
   in TSZ-treated PK-15 cells ([156]Fig. 6F), and the protein expression
   of CSFV N^pro obviously decreased with increasing doses of TSZ
   treatment ([157]Fig. 6G), suggesting that necroptosis may significantly
   reduce CSFV infection. To further confirm the limiting effect of
   necroptosis on CSFV infection, we treated CSFV-infected PK-15 cells
   with necrosulfonamide (NSA), an inhibitor of MLKL, and we could
   obviously observe that the inhibition of MLKL expression by NSA
   obviously promoted CSFV N^pro protein expression ([158]Fig. 6H). In
   PK-15 and 3D4/21 cells treated with both TSZ and rapamycin, the
   expression of CSFV N^pro drastically decreased with the addition of TSZ
   but was obviously restored with further rapamycin treatment ([159]Fig.
   6I). Taken together, these data strongly indicated that necroptosis is
   a limiting host factor for CSFV infection and that CSFV
   infection-induced autophagy was able to subvert this host defense
   mechanism to promote its sustained replication.

CSFV NS4A synergizes TRIM25 to induce mitophagy

   Previously, we showed that CSFV NS4A can promote the expression of
   TRIM25, which is likely to act as a key molecule linking CSFV NS4A and
   RIPK3. To gain insight into the biological function of TRIM25, we
   overexpressed TRIM25 in PK-15 and 3D4/21 cells, respectively, and
   measured the levels of p-mTOR, LC3-II, and P62. Overexpression of
   TRIM25 significantly increased LC3-II/LC3-I conversion and inhibited
   the expression of autophagy substrate P62 compared with the control,
   while the expression level of p-mTOR was not affected by TRIM25
   ([160]Fig. 7A). We further examined the effect of TRIM25 on the
   expression levels of mitochondria-associated proteins HSP60, TOMM20,
   COX IV, and VDAC. The results showed that the expression of
   mitochondria-associated proteins TOMM20, COX IV, and VDAC was
   significantly decreased in PK-15 cells overexpressing TRIM25 compared
   with the control, as was the expression of HSP60 and COX IV in 3D4/21
   cells ([161]Fig. 7B). To obtain a clearer and specific view of the
   cellular localization of TRIM25 to mitochondria and the effect of
   TRIM25 on mitochondrial number and morphology, we immunostained
   mitochondria with the mitochondrial probe Mito-Tracker, which allows us
   to observe that TRIM25 can partially localize to mitochondria and that
   PK-15 overexpressing TRIM25 appears granular to the mitochondria of
   3D4/21 cells, and mitochondria of cells transfected with empty vector
   appear filamentous ([162]Fig. 7C). The effect of TRIM25 on the
   formation and number of autophagosome-like vesicles was further
   examined by transmission electron microscopy. It was found that clear
   mitochondrial cristae (black arrows) could be observed mostly in the
   mitochondria of cells transfected with empty vector, while the
   mitochondrial cristae structure of PK-15 cells overexpressing TRIM25
   was disrupted, and some mitochondria were wrapped by double- or
   single-membrane structure, that is, typical mitophagosome structure
   appeared (red arrows), indicating that TRIM25 can perturb the
   mitochondrial metabolic activity of cells and lead to abnormal
   mitochondrial morphology ([163]Fig. 7D). In addition, the number of
   mitochondria encapsulated by autophagosome-like vesicles was
   significantly higher in PK-15 cells overexpressing TRIM25 than in empty
   vector-transfected cells. It is suggested that the TRIM25 protein can
   lead to the abnormality of cellular mitochondrial morphology and induce
   the occurrence of mitophagy.

Fig 7.

   [164]Fig 7
   [165]Open in a new tab

   CSFV NS4A synergizes TRIM25 to induce mitophagy. (A and B) PK-15 and
   3D4/21 cells were transfected with Myc-TRIM25 for 24 h, and pCMV-N1-Myc
   cells were transfected with control group. Expression levels of
   autophagy-associated proteins p-mTOR, LC3-II, and P62 and
   mitochondrial-associated proteins HSP60, TOMM20, COX IV, and VDAC in
   cell lysis products were detected by Western blot (left panel).
   Relative intensity related to internal parameters was analyzed using
   Image J plugin and plotted by GraphPad Prism 9 (right panel). Error
   bars indicate the mean (± SD) of three independent experiments. ns, P ≥
   0.05; **, P < 0.01; ***, P < 0.001 (two-way ANOVA). (C) PK-15 and
   3D4/21 cells were transfected with Myc-TRIM25 for 24 h, and pCMV-N1-Myc
   cells were set up as a control group. The cells were immunostained with
   Mito-Tracker, incubated with anti-Myc primary antibody, followed by
   Alexa fluor 488-conjugated anti-Rabbit-IgG secondary antibody (green),
   and the nuclei were stained with DAPI. Scale bars: 25 and 10 µm. (D)
   Mitochondrial structures and autophagy-like vesicles were visualized by
   transmission electron microscopy (left panel), and the number of
   autophagic vesicles with encapsulated mitochondria was quantified
   (right panel). Error bars indicate the mean (± SD) of five independent
   experiments. ***, P < 0.0005 (one-way ANOVA). (E) Myc-TRIM25 was
   co-transfected with pEGFP-C1 and GFP-NS4A in PK-15 cells, respectively,
   while a control group of cells was set up in which pCMV-N1-Myc was
   co-transfected with pEGFP-C1 and GFP-NS4A, respectively. Expression
   levels of autophagy-associated proteins p-mTOR, LC3-II, and P62 and
   mitochondrial-associated proteins HSP60, TOMM20, COX IV, and VDAC in
   cell lysis products were detected by Western blot. (F) The relative
   intensity of the detection indexes in (E), related to the internal
   control GAPDH, was analyzed with Image J plugin and plotted by GraphPad
   Prism 9. Error bars indicate the mean (± SD) of three independent
   experiments. *, P < 0.1; and ****, P < 0.0001 (two-way ANOVA). (G) Si
   TRIM25 was co-transfected with pEGFP-C1 and GFP-NS4A in PK-15 cells,
   respectively, while a control group of cells was set up in which Si NC
   was co-transfected with pEGFP-C1 and GFP-NS4A, respectively. Expression
   levels of autophagy-associated proteins p-mTOR, LC3-II, and P62 and
   mitochondrial-associated proteins HSP60, TOMM20, COX IV, and VDAC in
   cell lysis products were detected by Western blot. (H) The relative
   intensity of the detection indexes in (G), related to the internal
   control GAPDH, was analyzed with Image J plugin and plotted by GraphPad
   Prism 9. Error bars indicate the mean (± SD) of three independent
   experiments. ns, P ≥ 0.05 and ****, P < 0.0001 (two-way ANOVA). (I)
   HA-NS4A was co-transfected with Myc-TRIM25 and Si TRIM25 into PK-15
   cells, respectively, and a cell control transfected with HA-NS4A only
   and pCAGGs-HA only was set up, along with dual-fluorescent reporter
   plasmid mRFP-GFP-LC3 to visualize the autophagic flux. (J) HA-NS4A was
   co-transfected with Myc-TRIM25 and Si TRIM25 into PK-15 cells,
   respectively, and a cell control transfected with HA-NS4A only and
   pCAGGs-HA only was set up, along with mitophagy dual-fluorescent
   reporter plasmid Mito-mRFP-EGFP to visualize the mitochondria-lysosome
   delivery. The fluorescence intensity of GFP-Mito (green) and RFP-Mito
   (red) was analyzed in Fig. S4E with Image J plugin and plotted by
   GraphPad Prism 9. Error bars indicate the mean (± SD) of at least three
   independent experiments. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001
   (two-way ANOVA).

   Consistent with this previous report, CSFV NS4A was also able to
   partially colocalize with mitochondria; however, the knockdown of
   TRIM25 was able to suppress the expression of CSFV NS4A and inhibit its
   degree of colocalization with mitochondria (Fig. S4D). CSFV NS4A, as an
   important non-structural protein of CSFV, forms a
   replication-associated complex with NS3, NS4B, NS5A, and NS5B to
   participate in viral replication. At present, little is known about the
   role of CSFV NS4A protein in the process of mitophagy. We found that
   overexpression of CSFV NS4A promoted the degradation of autophagy
   substrate P62 as well as the conversion of LC3-II/LC3-I, and more
   importantly, its ability to inhibit the expression of
   mitochondria-related proteins ([166]Fig. 7E), and a higher number of
   mitochondria encapsulated by autophagosome-like vesicles in NS4A
   overexpressing PK-15 cells than in empty vector-transfected cells were
   observed by transmission electron microscopy (Fig. S4A). Even more,
   overexpression of TRIM25 was able to deepen the effect of NS4A on the
   degradation of mitochondria-related proteins in PK-15 ([167]Fig. 7E and
   F) and 3D4/21 (Fig. S4B) cells and the induction of autophagy.
   Conversely, the knockdown of TRIM25 was able to specifically restore
   the inhibitory effect of NS4A on mitochondria-related proteins in PK-15
   ([168]Fig. 7G and H) and 3D4/21 (Fig. S4C) cells and attenuate the
   induction effect on autophagy. We co-transfected recombinant plasmid
   HA-NS4A expressing CSFV NS4A with Myc-TRIM25 and Si TRIM25,
   respectively, into PK-15, to monitor autophagic flux using
   mRFP-GFP-LC3, and strong aggregation of LC3 with orange fluorescence
   was observed in NS4A overexpressing, especially PK-15 overexpressing
   both NS4A and TRIM 25, whereas it was reduced upon siRNA-mediated
   knockdown of TRIM 25 ([169]Fig. 7I). We further employed the mitophagy
   dual fluorescent reporter plasmid Mito-mRFP-EGFP to observe the
   delivery of mitochondria-lysosomes in above co-transfected HA-NS4A and
   Myc-TRIM25/Si TRIM25 cells. The results showed that the mitochondria of
   PK-15 ([170]Fig. 7J and S4E) and 3D4/21 (Fig. S4F) cells overexpressing
   both NS4A and TRIM25 exhibited orange-yellow fluorescence, whereas the
   mitochondria of cells transfected with empty vector exhibited yellow
   fluorescence, and the mitochondria of cells transfected with both NS4A
   and Si TRIM25 exhibited greenish yellow fluorescence. The progress of
   autophagy/mitophagy was judged according to the quenching degree of
   green fluorescence, indicating that CSFV NS4A cooperated with TRIM25 to
   induce autophagy, causing the degradation of mitochondria by cell
   lysosomes.

RIPK3 localizes to the outer mitochondrial membrane and is degraded by
TRIM25-induced mitophagy under CSFV infection

   Previously, we found that CSFV NS4A specifically mediates the
   degradation of RIPK3 via TRIM25, the specific molecular mechanism of
   which remains unclear. Currently, there are mainly two main
   intracellular protein degradation systems, the ubiquitin-proteasome
   system (UPS) and autophagy-lysosome pathway ([171]47), we further
   explored which degradation system TRIM25 mainly employs to regulate the
   degradation of RIPK3 after CSFV infection. To this end, CSFV-infected
   PK-15 cells were co-transfected with plasmids expressing RIPK3 and
   TRIM25, and then the cells were treated with the protease inhibitor
   MG132, rapamycin, and the autophagy inhibitor bafilomycin A1 (Baf A1),
   respectively. The results showed that both MG132 and Baf A1 could
   increase the protein expression level of RIPK3 ([172]Fig. 8A).
   Consistent with previous reports, we found that TRIM25 was able to
   ubiquitinate RIPK3 after CSFV infection to mediate its degradation via
   the UPS ([173]Fig. 8B). Co-transfection of HA-RIPK3 and HA-MLKL
   separately with GFP-LC3 into PK-15 cells allowed to observe that RIPK3,
   but not MLKL, could co-localize with LC3 ([174]Fig. 8C and 5A).
   Furthermore, mitochondria in RIPK3-overexpressed PK-15 cells were
   immunostained with Mito-tracker, a mitochondrial probe, to observe the
   spatial distribution of RIPK3 with mitochondria. Interestingly, RIPK3
   was able to wrap around the outer mitochondrial membrane, and the
   degree of co-localization of RIPK3 with mitochondria was obviously
   enhanced in the knockdown of TRIM25 ([175]Fig. 8D). In contrast, there
   was no significant colocalization of MLKL with mitochondria in either
   the presence or absence of Si TRIM25 (Fig. S5B). In addition, we
   labeled the lysosomes of PK-15 cells co-expressing RIPK3 and LC3 with
   the lysosomal outer membrane protein CD63 under CSFV infection
   conditions and could find that RIPK3 was able to co-localize with LC3
   to lysosomes, and the intensity of co-localization with LC3 was reduced
   after knockdown of TRIM25 ([176]Fig. 8E).

Fig 8.

   [177]Fig 8
   [178]Open in a new tab

   RIPK3 is able to be degraded by TRIM25-induced mitophagy under CSFV
   infection. (A) PK-15 cells infected with CSFV (MOI = 1) were
   co-transfected with Myc-TRIM25 and HA-RIPK3 for 24 h and then treated
   with MG132 (10 mM), rapamycin (100 nM), or Baf A1 (100 nM) for 6 h, and
   cell lysates were detected by Western blot (left panel). Relative
   intensity related to internal control GAPDH was analyzed using Image J
   plugin and plotted by GraphPad Prism 9 (right panel). Error bars
   indicate the mean (± SD) of three independent experiments. ns, P ≥ 0.05
   and ****, P < 0.0001 (one-way ANOVA). (B) PK-15 cells infected with
   CSFV (MOI = 1) were co-transfected with Myc-TRIM25, Flag-RIPK3, and
   His-Ub for 24 h and then treated with MG132 (10 mM) for 6 h, and the
   cells were lysed and used for IP experiments. (C) HA-RIPK3 and GFP-LC3
   were co-transfected into PK-15 cells for 24 h and then treated with
   rapamycin (100 nM) for 6 h while setting up a control group of cells
   treated with dimethyl sulfoxide (DMSO). The cells were incubated with
   anti-HA primary antibody, followed by Alexa fluor 555-conjugated
   anti-Rabbit-IgG secondary antibody (red), and the nuclei were stained
   with DAPI (left panel). Scale bars: 100 and 25 µm. The fluorescence
   intensity profiles of GFP-LC3 (green) and HA-RIPK3 (red) were analyzed
   using Image J plugin and plotted by GraphPad Prism 9 (right panel). (D)
   PK-15 cells were co-transfected with HA-RIPK3 together with Si TRIM25
   for 24 h while setting up a control group of co-transfected Si NC
   cells, the cells were immunostained with Mito-Tracker, incubated with
   anti-HA primary antibody, followed by Alexa fluor 488-conjugated
   anti-Mouse-IgG secondary antibody (green), and nuclei were stained with
   DAPI (left panel). Scale bars: 100 and 10 µm. The fluorescence
   intensity profiles of HA-RIPK3 (green) and Mito-Tracker (red) were
   analyzed using Image J plugin and plotted by GraphPad Prism 9 (middle
   panel). Pearson’s correlation and overlap co-efficient of HA-RIPK3 and
   Mito-Tracker were analyzed using Image J plugin and plotted by GraphPad
   Prism 9 (right panel). Error bars indicate the mean (± SD) of at least
   three independent experiments. ns, P ≥ 0.05 and *, P < 0.1 (one-way
   ANOVA). (E) PK-15 cells infected with CSFV (MOI = 1) were
   co-transfected with Si TRIM25, Flag-RIPK3, and GFP-LC3 for 24 h while
   setting up a control group of co-transfected Si NC cells, the fixed
   cells were incubated with anti-Flag primary antibody and anti-CD63
   primary antibody, respectively, and then stained with Alexa fluor
   555-conjugated anti-Mouse-IgG secondary antibody (red) and Dylight
   405-labeled anti-Rabbit-IgG secondary antibody (blue) (left panel).
   Scale bars: 10 µm. The fluorescence intensity profiles of Flag-RIPK3
   (blue), GFP-LC3(green), and CD63(red) were analyzed using Image J
   plugin and plotted by GraphPad Prism 9 (middle panel). Pearson’s
   correlation and overlap co-efficient were analyzed using Image J plugin
   and plotted by GraphPad Prism 9 (right panel). Error bars indicate the
   mean (± SD) of at least three independent experiments. ***, P < 0.001
   and ****, P < 0.0001 (one-way ANOVA). (F) The mRNA expressions of
   TRIM25 in the tonsil, lung, lymph node, spleen, kidney, and thymus of
   CSFV-infected and non-infected piglets were determined by qRT-PCR. The
   fold changes were related to the internal control β-actin. Error bars
   indicate the mean (± SD) of three independent experiments. ****, P <
   0.0001 (two-way ANOVA). (G) To establish PK-15 (upper panel) and 3D4/21
   (lower panel) cell models of CSFV infection in vitro, cells were
   collected at 24 and 48 h post-infection, and protein expression levels
   of TRIM25 and CSFV N^pro were detected by Western blot. (H) PK-15 (left
   panel) and 3D4/21 (right panel) cells infected with CSFV (MOI = 1) were
   transfected with Myc-TRIM25 or Si TRIM25 for 48 h, and pCMV-N1-Myc or
   Si NC cells were transfected with a control group. The cell
   supernatants were collected for TCID[50] assay. Error bars indicate the
   mean (± SD) of three independent experiments. *, P < 0.1; ***, P <
   0.001; ****, P < 0.0001 (one-way ANOVA). (I) PK-15 (left panel) and
   3D4/21 (right panel) cells infected with CSFV (MOI = 1) were
   transfected with Myc-TRIM25 for 24 and 48 h, and pCMV-N1-Myc cells were
   transfected with a control group. Expression levels of TRIM25 and CSFV
   E2/N^pro in cell lysis products were detected by Western blot.

CSFV promotes the expression of TRIM25, which enhances CSFV replication

   TRIM25, as a key molecule linking CSFV NS4A and TRIM25, how is its
   interaction with CSFV? We examined the mRNA levels of TRIM25 in
   different tissues including lung, kidney, and lymph nodes in
   CSFV-uninfected and -infected piglets and found that CSFV promotes
   TRIM25 transcription in different tissues, with distinct tissue tropism
   for TRIM25 ([179]Fig. 8F). Western blot results showed that CSFV in
   vitro infection of PK-15 and 3D4/2 cells for 24 and 48 h was able to
   obviously promote the protein expression of TRIM25 ([180]Fig. 8G). To
   further explore the role of TRIM25 in CSFV infection, a TRIM25
   expression plasmid was transfected into PK-15 and 3D4/21 cells,
   respectively, followed by CSFV infection. Immunoblotting showed that
   the expression of CSFV E2 protein was significantly increased in PK-15
   cells overexpressing TRIM25 compared with the control, and the
   expression of CSFV N^pro protein was also significantly increased in
   3D4/21 cells overexpressing TRIM25 compared with the control ([181]Fig.
   8I). We knocked down TRIM25 expression using specific interfering RNA
   targeting or transfected TRIM25 expression plasmids into PK-15 and
   3D4/21 cells, followed by detection of viral titers in the cell
   supernatant after 48 h of CSFV infection. The results showed that
   functional silencing of TRIM25 significantly suppressed CSFV titers in
   cell supernatants, whereas overexpression of TRIM25 promoted increased
   CSFV infectivity titers ([182]Fig. 8H).

CSFV promotes RIPK3 autophagic degradation through the binding of autophagy
receptor NDP52

   Selective autophagy requires the engagement of autophagy receptors,
   which specifically recognize and traffic autophagy substrates, thereby
   targeting the degradation of autophagy substrates, endowing the
   material dependent on autophagic degradation under extremely
   sophisticated dynamic regulation ([183]22, [184]25). We sought to
   identify the cargo receptor involved in the autophagic degradation of
   RIPK3, and the Co-IP results showed that RIPK3 was able to interact
   with P62 and NDP52, and overexpression of NDP52 was able to obviously
   inhibit the expression of RIPK3 compared with other cargo receptors
   ([185]Fig. 9A). Consistent with this result, we were able to observe a
   clear co-localization of NDP52 with RIPK3 in the PK-15 cytoplasm
   ([186]Fig. 9B). Notably, the distribution of RIPK3 protein expressed
   alone was more diffuse in cells, whereas in cells simultaneously
   expressing both proteins, the distribution of RIPK3 was more clustered
   and highly coincident with NDP52 ([187]Fig. 9B). In addition, Baf A1
   treatment of cells was able to more prominently observe the
   colocalization of RIPK3, LC3, and NDP52 in CSFV-infected PK-15 cells
   ([188]Fig. 9C). Furthermore, significant co-localization of NS4A with
   NDP52 (Fig. S3B) was observed during CSFV infection. We speculate that
   NDP52 may play an important role in regulating the autophagic
   degradation of RIPK3 in CSFV-infected cells.

Fig 9.

   [189]Fig 9
   [190]Open in a new tab

   The autophagy receptor NDP52 binds to RIPK3 and promotes its autophagic
   degradation. (A) HEK-293T cells were co-transfected with HA-RIPK3
   together with pCMV-3× Flag, Flag-P62, Flag-NDP52, Flag-NBR1, or
   Flag-TAX1BP1 for 24 h, the cells were lysed and used for Co-IP
   experiments, and cell lysates were detected by Western blot. (B)
   Flag-NDP52 and HA-RIPK3 were co-transfected into PK-15 cells for 24 h
   while setting up a control group of cells transfected with pCMV-3×
   Flag. The fixed cells were incubated with anti-Flag primary antibody
   and anti-HA primary antibody, followed by Alexa fluor 488-conjugated
   anti-Mouse-IgG secondary antibody (green) and Alexa fluor
   555-conjugated anti-Rabbit-IgG secondary antibody (red), and the nuclei
   were stained with DAPI. Scale bars: 100 and 10 µm. (C) PK-15 cells
   infected with CSFV (MOI = 1) were co-transfected with Flag-NDP52,
   GFP-LC3, and HA-RIPK3 for 24 h and then treated with Baf A1 (100 nM)
   for 6 h while setting up a control group of cells treated with dimethyl
   sulfoxide, the fixed cells were incubated with anti-Flag primary
   antibody and anti-HA primary antibody, respectively, and then stained
   with Alexa fluor 555-conjugated anti-Rabbit-IgG secondary antibody
   (red) and Dylight 405-labeled anti-Mouse-IgG secondary antibody (blue).
   Scale bars: 10 µm. (D) Schematic of RIPK3 domain structure and
   deletion/mutation structure. (E) PK-15 cells were co-transfected with
   Flag-NDP52 together with pCAGGs-HA, Flag-RIPK3 WT, Flag-RIPK3 mutRHIM,
   Flag-RIPK3 ΔCD, or Flag-RIPK3 ΔKD for 24 h, the cells were lysed and
   used for Co-IP experiments, and cell lysates were detected by Western
   blot. (F) HEK 293T cells were co-transfected with Flag-NDP52 together
   with pCAGGs-HA, Flag-RIPK3 WT, Flag-RIPK3 mutRHIM, Flag-RIPK3 ΔCD, or
   Flag-RIPK3 ΔKD for 24 h, the fixed cells were incubated with anti-Flag
   primary antibody and anti-HA primary antibody, followed by Alexa fluor
   488-conjugated anti-Mouse-IgG secondary antibody (green) and Alexa
   fluor 555-conjugated anti-Rabbit-IgG secondary antibody (red), and the
   nuclei were stained with DAPI. Scale bars: 100 and 10 µm. (G) Proposed
   a model by which CSFV infection in vitro limits necroptosis via the
   autophagy pathway. In vitro, the induction of autophagy by CSFV at a
   later stage of infection clearly restricts necroptosis. Mechanistic
   studies revealed that CSFV NS4A protein promoted TRIM25 expression,
   synergistically induced the occurrence of mitophagy, targeted the
   autophagic degradation of RIPK3 to block the progression of necroptosis
   occurrence, and achieved persistent viral infection. Interestingly, we
   found that RIPK3 was able to specifically localize at the outer
   mitochondrial membrane, and the autophagy receptor NDP52 was most
   likely involved in the autophagic degradation of RIPK3 during CSFV
   infection.

   To determine which domain of NDP52 interacts with RIPK3, we constructed
   constructs for truncation of the C-terminal domain (ΔCD, aa 1–299) and
   kinase domain (ΔKD, aa 300–478) of a truncation plasmid of RIPK3. In
   addition to mutating the RHIM (aa 453–470), the mutRHIM of RIPK3 was
   generated by overlap extension PCR, changing “VQIG” from 460 to 463 to
   “AAAA” ([191]Fig. 9D). The results showed that NDP52 was able to
   interact with all three mutant or truncated forms of RIPK3 ([192]Fig.
   9E and F). Autophagy receptor proteins, on one hand, were able to
   recognize and bind autophagy substrates through a ubiquitin-associated
   (UBA) domain-dependent or -independent manner while anchored on the
   autophagosome membrane through their LIRs, and subsequently, the
   autophagosome and lysosome fuse, and the substrates are degraded in the
   lysosome ([193]26, [194]48, [195]49). NDP52, as a UBA-containing
   receptor, can bind ubiquitinated autophagy substrates, and the ability
   of NDP52 to interact with different mutations or truncations of RIPK3
   may be related to its presence of multiple ubiquitination sites across
   domains.

DISCUSSION

   The occurrence and development of infectious diseases are essentially a
   process of continuous mutual games between pathogenic microorganisms
   and their hosts. Viruses, as intracellular hosts, long-term “struggle”
   with the host during the evolution process, leading to the emergence of
   a highly complex and elaborate anti-infective immune system and
   mechanisms in the body; meanwhile, the virus also acquires various
   mechanisms to escape the body’s immune response, which can lead to a
   worse response.

   CSFV has a strong tropism for epithelial cells and immune cells and is
   a typical immunosuppressive virus that seriously damages the
   hematopoietic and immune systems. Viremia can occur early in CSFV
   infection, followed by immunopathological features of lymphopenia
   including B lymphocytes, cytotoxic T cells, and helper T cells, and
   lymphopenia is the leading cause of immunosuppression. Previously, it
   was shown that high titers of CSFV can be detected in bone marrow at an
   early stage of viral infection, and peripheral T lymphopenia is closely
   related to the apoptotic damage of bone marrow lymphocytes caused by
   CSFV ([196]50). Further studies found that granulocytopenia and
   disordered bone marrow function were the results of hematopoietic cell
   apoptosis due to the interplay of viral and host mechanisms ([197]51).
   Autophagy is generally considered as a cell survival mechanism and is
   closely related to apoptosis ([198]52, [199]53). During CSFV in vivo
   infection, autophagy leads to the death of tissue cells in the immune
   organs of infected pigs and mediates the reduction of T-zone
   lymphocytes ([200]54). However, a previous study found that instead of
   clearing CSFV from infected cells, the autophagy pathway was utilized
   by CSFV to provide favorable conditions for its replication in cells
   ([201]55). The pathogenesis features resulting from CSFV infection are
   similar to those caused by a “cytokine storm,” i.e., the disorder of
   massive secretion of inflammation-related cytokines is closely related
   to disease progression ([202]56). Necroptosis distinguishes itself from
   apoptosis by its pro-inflammatory role, necroptosis with rupture of the
   cell membrane, accompanied by content release, and in particular, the
   inflammatory factors released initiate a cascade of inflammatory
   responses. Earlier studies have long considered apoptosis as the main
   mechanism underlying the lymphopenia syndrome upon CSFV infection.
   However, an increased number of cells with distinct necrotic features
   in PBMCs and spleens of peripheral immune organs of CSFV-infected
   piglets were observed in the present study, which were accompanied by
   the inflammatory-related factors TNF-α and IL-1β, the expression of
   necroptosis executioner MLKL was elevated in PBMC cells from CSFV
   infected piglets, and IHC also further validated the high expression of
   necroptosis marker proteins RIPK1, RIPK3, and MLKL in the spleen of
   CSFV-infected piglets. Showed that CSFV infection induced necroptosis
   occurrence in PBMCs and spleen in vivo.

   Cell death is an essential component of the host’s immune response
   against invading microbial pathogens. However, it is not surprising for
   the fact that some viruses can balance host defenses against infection
   ([203]57, [204]58). Many viruses with non-cytopathic effects, including
   CSFV, can hijack various mechanisms to relieve ongoing interference
   from host cells to remain infected. In our further in vitro assays, it
   was not difficult to find that CSFV infection of PK-15 and 3D4/21 cells
   could induce necroptosis at the early stage, while it inhibited at the
   middle and late stages, which might also be an important reason why
   CSFV infection in vitro could not cause cytopathic effects. A similar
   phenomenon occurs in the context of murine cytomegalovirus infection, a
   virus that inhibits necroptosis of murine cells by disrupting
   RHIM-dependent signal transduction ([205]15). In addition, human
   cytomegalovirus has also been shown to prevent necrotic cell death in
   human fibroblasts and monocytes ([206]28, [207]59). α-Members of the
   herpesvirus subfamily, HSV-1 and herpes simplex virus type 2, exhibit
   opposing activities in manipulating necroptosis, depending on the host
   species ([208]60). CSFV induces necroptosis in PBMCs and spleen in vivo
   but inhibits necroptosis levels in porcine kidney-derived cell line
   (PK-15) and alveolar macrophage cell line (3D4/21) in vitro at 60 dpi.
   On one hand, the induction of necroptosis is closely related to the
   levels of autophagy induced during CSFV infection, suggesting that
   autophagy induction may precede necroptosis, ultimately leading to the
   suppression of necroptosis levels. On the other hand, CSFV infection
   displays different tissue tropisms and induces varying degrees of
   pathology in different tissues. For example, it can cause splenic
   marginal hemorrhage and infarction as well as dense subcapsular
   petechiae in the kidneys. However, CSFV infection in vitro does not
   induce cellular pathology. This suggests that the activation/inhibition
   of necroptosis by CSFV in vivo and in vitro is of great relevance to
   the corresponding tissue/target cell types.

   The process of programmed necrosis is accompanied by autophagy, and
   programmed necrosis can activate autophagy, whereas its induced
   activation of the necrosome complex can in turn occur downstream of
   autophagosome formation ([209]61). This illustrates that there is a
   complex link between programmed necrosis and autophagy that may
   simultaneously occur during cell death ([210]28). CSFV was able to
   reach the peak of viral replication at 48 h of infection in vitro,
   accompanied by the induction of a strong autophagic response. Rather
   paradoxically, the necroptosis process at this point is progressively
   inhibited with a continued extension of the CSFV infection time, which
   is known to be long, and it is reasonable to speculate that there is
   some corollary link between these two modes of cell death in CSFV
   infection conditions. As expected, we treated CSFV-infected cells with
   an autophagy/mitophagy agonist, and unsurprisingly, the level of
   necroptosis induced at the initial stage of CSFV infection was markedly
   suppressed. The observation that RIPK3, but not MLKL, was able to
   localize to the outer mitochondrial membrane and specifically
   co-localizes with LC3 further explains the outcome of necroptosis
   inhibition by treatment with mitophagy agonists. Importantly, we found
   that both necroptosis inducer TSZ treatment and overexpression of RIPK3
   and MLKL significantly inhibited CSFV replication and infection
   progression, whereas the necroptosis executioner MLKL clearly promoted
   CSFV replication when inhibited by NSA, suggesting that necroptosis is
   a limiting host factor for CSFV infection and that CSFV
   infection-induced autophagy disintegrates this host defense mechanism
   to promote its sustained replication.

   To further explore the molecular mechanism by which CSFV infection
   inhibits necroptosis in vitro, we found that CSFV NS4A protein was able
   to interact with the E3 ubiquitin ligase TRIM25, which was first
   identified ubiquitously because of its antiviral role, and
   TRIM25-mediated ubiquitination activation of RIG-I is an important step
   in initiating intracellular antiviral responses ([211]62). In addition,
   TRIM25 can regulate host antiviral capacity through other pathways,
   including negative regulation of RIG-I, activation of melanoma
   differentiation-associated gene 5, enhancing the activity of
   zinc-finger antiviral proteins (ZAP), and inhibiting viral RNA
   synthesis ([212]63, [213]64). Interestingly, we found in our study that
   CSFV NS4A protein obviously promoted the expression of trim25 in a
   dose-dependent manner, while TRIM25 did not play an antiviral role as
   it did in other virus infections, unexpectedly obviously promoted the
   infection progress of CSFV, correspondingly, TRIM25 was also able to
   stabilize the expression of NS4A protein. Our further study revealed
   that NS4A was able to form a complex with TRIM25, RIPK3, following a
   previous report that TRIM25 was a novel “brake” molecule targeting the
   necroptotic key protein RIPK3 and was able to modify RIPK3 by
   ubiquitination to inhibit TNF-induced cell death. We verified that CSFV
   NS4A could suppress the expression of RIPK3 via TRIM25 functional
   silencing, and more importantly, CSFV NS4A induced mitochondrial
   fission and mitophagy with the assistance of TRIM25, suggesting that
   TRIM25 is likely to serve as a specific and critical target for CSFV
   infection escape. Based on our finding that TRIM25 mediates RIPK3
   degradation through both the UPS and autophagy-lysosome pathways under
   CSFV infection conditions, a biological phenomenon somewhat
   contradicted by the finding that TRIM25 mainly degrades RIPK3 through
   the ubiquitination pathway, we speculated that CSFV engagement might
   alter the RIPK3 degradation pathway to some extent. In addition, it has
   been mentioned in other reports that RIPK3 was able to be ubiquitinated
   and then undergo lysosome-dependent degradation ([214]65, [215]66),
   consistent with our finding of enhanced colocalization of RIPK3 with
   lysosomes after CSFV infection.

   Autophagic degradation of substrates requires selective recognition and
   trafficking of cargo receptors, and we screened for evidence that
   autophagic degradation of RIPK3 may require, but is not limited to, the
   involvement of NDP52, but we did not find a specific site where NDP52
   interacts with RIPK3. As in our study, RIPK3 with truncated C-terminal
   domain, kinase domain, or mutated RHIM region was still able to
   interact with NDP52, we hypothesized that NDP52 acts as an autophagy
   receptor containing a UBA and depends on ubiquitin-binding substrates
   for ubiquitination, whereas porcine RIPK3, similar to human RIPK3,
   contains multiple ubiquitination sites spanning domains. In subsequent
   studies, we could sequentially construct RIPK3 variants with serial
   ubiquitination site mutations, further confirming the binding site of
   NDP52. In addition, the association of TRIM25 with NDP52 is still
   missing in this study, and it remains to be confirmed that TRIM25
   mediates autophagic degradation of RIPK3 via the NDP52 receptor under
   CSFV infection, and the relationship axis remains to be refined.

   Collectively, we found that CSFV infection in vivo was able to activate
   necroptosis in PBMCs and spleen, leading to lymphocyte necrosis.
   Whereas, CSFV infection markedly inhibited necroptosis by inducing
   autophagy in vitro. It was further found that the CSFV NS4A protein
   interacts with the E3 ubiquitin ligase TRIM25 and is targeted to
   mediate the autophagic degradation of RIPK3, thereby blocking the
   progression of necroptosis to achieve persistent viral infection
   ([216]Fig. 9G). Our findings shed further light on the mechanisms that
   cause T lymphocyte exhaustion during CSFV infection, provide evidence
   supporting an important role for CSFV in counteracting host cell
   necrosis, and enrich our knowledge of pathogens that evade this host
   defense.

MATERIALS AND METHODS

Cell and virus

   The swine kidney cell line PK-15 (ATCC, CCL-33) and HEK293T (ATCC,
   CRL-1573) cells were grown in DMEM (Thermo Fisher Scientific, 11995500)
   supplemented with 10% (vol/vol) fetal bovine serum (FBS, Thermo Fisher
   Scientific, 10099), and macrophage cell line 3D4/21 (ATCC, CRL-2845)
   cells were maintained in complete RPMI 1640 medium (Thermo Fisher
   Scientific, 1175093) containing 10% FBS. Cells were cultured at 37°C in
   a 5% CO[2] incubator. The CSFV-Shimen strain was stocked in our
   laboratory and proliferated in the cultured PK-15 cells. Virus titers
   were determined with 50% tissue culture infective dose (TCID[50])
   assays.

Animal experiment

   All procedures were conducted following the regulations of the
   Laboratory Animal Ethics Committee of South China Agricultural
   University. Ten 2-month-old Tibetan pigs negative for the porcine
   reproductive and respiratory syndrome, pseudorabies virus, and porcine
   parvovirus, purchased from the Experimental Animal Center of Southern
   Medical University, were randomly divided into two groups, one of which
   was intramuscularly injected into the bilateral neck with 10^5 TCID[50]
   of CSFV and the other with the same volume of phosphate-buffered saline
   (PBS) ([217]42, [218]67). Before or after CSFV infection, the anterior
   vena cava blood of piglets was sterilely collected into heparin sodium
   anticoagulant tubes every other day, and peripheral blood mononuclear
   cells were isolated and subjected to flow cytometry analysis. Then
   CD3^+ T cells in PBMCs and splenic lymphocytes isolated at 7 d
   post-infection (dpi) were used for transcriptomic analysis. Piglets
   were euthanized at 7 dpi, immune tissues (tonsil, thymus, spleen, and
   lymph node) and non-immune tissues (lung and kidney) were collected and
   used for hematoxylin-eosin staining, quantitative reverse transcription
   polymerase chain reaction, Western blot, and IHC.

Reagents and antibodies

   The chemical reagents used in this study are as follows: dimethyl
   sulfoxide (Sigma-Aldrich, V900090), rapamycin (Cell Signaling
   Technology, 9904), bafilomycin A1 (MedChemExpress, HY-100558), carbonyl
   cyanide 3-chlorophenylhydrazone (Sigma-Aldrich, C2759), recombinant
   TNF-α protein (Abcam, ab9642), SM-164 (Selleck, S7089), Z-VAD-FMK
   (MedChemExpress, HY-16658B), and necrosulfonamide (MedChemExpress,
   HY-100573). Opti-MEM I reduced serum medium (31985088) and
   Lipofectamine 3000 transfection reagent (L3000015) were purchased from
   Thermo Fisher Scientific.

   The primary antibodies used in this study were as follows: mouse
   monoclonal anti-TNF-α (Santa Cruz Biotechnology, sc-52746), mouse
   monoclonal anti-IL-1β (Santa Cruz Biotechnology, sc-32294), rabbit
   monoclonal anti-TRIM25 (Abcam, ab167154), rabbit polyclonal anti-RIPK1
   (Beyotime Biotechnology, AF7896), mouse monoclonal anti-RIPK3 (Santa
   Cruz Biotechnology, sc-374639), rabbit monoclonal anti-MLKL (Santa Cruz
   Biotechnology, sc-293201), mouse monoclonal anti-ZBP1 (Santa Cruz
   Biotechnology, sc-166344), rabbit monoclonal anti-p62 (Cell Signaling
   Technology, 8025), rabbit polyclonal anti-LC3B (Cell Signaling
   Technology, 2775), rabbit monoclonal anti-VDAC1 (Beyotime
   Biotechnology, AF-1024), rabbit monoclonal anti-TOMM20 (Beyotime
   Biotechnology, AF-1717), rabbit polyclonal anti-COX IV (Beyotime
   Biotechnology, AC610), rabbit monoclonal anti-HSP60 (Beyotime
   Biotechnology, AG2237), mouse monoclonal anti-ubiquitin (Cell Signaling
   Technology, 3936), mouse monoclonal anti-CSFV E2 (Jai Balajee Trading
   Company, 9011), rabbit polyclonal anti-N^pro (kindly donated by
   professor Xinglong Yu, Hunan Agricultural University, China), mouse
   monoclonal anti-GAPDH (Beyotime Biotechnology, AG019), and mouse
   monoclonal anti-Tubulin (Beyotime Biotechnology, AT819). The secondary
   antibodies used for immunofluorescence were Alexa Fluor 555 donkey
   anti-rabbit IgG (Beyotime Biotechnology, A0412), Alexa Fluor 488 goat
   anti-mouse IgG (Beyotime Biotechnology, A0428), Dylight 405 goat
   anti-rabbit IgG (Beyotime Biotechnology, A0605), and Alexa Fluor 555
   donkey anti-mouse IgG (Beyotime Biotechnology, A0460).

Transcriptomic analysis

   Total RNA was extracted from CD3^+ T cells using Trizol reagent. The
   concentration and integrity of RNA were assessed by assays using K550
   and Agilent 2200 Tapestation, respectively. After qualifying samples,
   eukaryotic mRNA was enriched using magnetic beads with Oligo (dT);
   after interrupting the mRNA, the mRNA was used as a template to
   construct a complementary DNA (cDNA) library. After the construction,
   the insert length and effective concentration of the library were
   checked to ensure the quality of the library. After the library
   examination was qualified, the different libraries were pooled
   according to the amount of under-target data for sequencing using the
   Illumina platform. To guarantee the quality of subsequent analysis,
   adaptor sequences need to be removed, and reads with low quality (low
   quality, bases with a base quality value less than or equal to 25
   accounted for more than 60% of the whole reads) and N (N means base
   information cannot be determined) proportion greater than 5% were
   filtered out to obtain clean reads that could be used for subsequent
   analysis. Clean reads were sequence aligned to the indicated genome
   using HISAT2 software, obtaining information on their position on the
   reference genome.

   Fragments per kilobase of exon model per million mapped reads were
   calculated for each gene in individual samples using FeatureCounts, a
   method that eliminates the impact of differences in length and
   sequencing on calculated gene expression, i.e., normalizing for gene
   expression. Differential expression analysis among genes was performed
   using DESeq2 software, and differentially expressed genes (DEGs) were
   screened with the criteria of |log2foldchange| >1 and P [adj] <0.05.
   The functional annotation of DEGs by gene oncology
   ([219]http://www.geneontology.org) and the enrichment analysis by Kyoto
   Encyclopedia of Genes and Genomes ([220]http://www.genome.jp/kegg/)
   pathway were performed using DESeq2 software with a hypergeometric test
   to discover the biological processes and signal transduction pathways
   that DEGs were involved in.

qRT-PCR assay

   The qRT-PCR assay used in this study was performed as previously
   described ([221]42). Briefly, total RNA was extracted from the cells of
   each group using TRIzol reagent (Invitrogen, 15596026) or a total RNA
   Kit I (Omega, R6834-02). cDNA was synthesized using HiScript II Q RT
   SuperMix (Vazyme, R223-01) for detecting relative mRNA expression of
   the indicated genes. The qRT- PCR assay was performed using ChamQ
   Universal SYBR qPCR Master Mix (Vazyme, Q711-02) using an iQ5 iCycler
   detection system (Bio-Rad, 1855200). Relative mRNA expressions of the
   specific genes were determined using the 2^−ΔΔCT method and normalized
   to the housekeeping gene β-actin.

Flow cytometry assay

   The collected PBMCs were washed in PBS and incubated for 15 min with
   FITC mouse anti-Pig CD3 (BD Biosciences, 559582) and PE mouse anti-Pig
   CD4a (BD Biosciences, 559586), according to the manufacturer’s
   protocol. PK-15 cells after a viral infection or drug treatment were
   collected using trypsin-EDTA and washed in PBS. The cells were
   resuspended in annexin buffer and incubated for 15 min with 2.5  µg/mL
   PI (BD Biosciences) and Annexin V (BD Biosciences), according to the
   manufacturer’s protocol. The antibody-labeled or necrotic cells were
   analyzed using the flow cytometer (Beckman Coulter Life Science,
   CytoFLEX). Flow cytometry analyses were reproduced by three independent
   experiments.

Western blot analysis

   Cells collected from each group were washed twice in cold PBS and then
   lysed on ice with cell lysis buffer (Beyotime Biotechnology, P0013B)
   containing 1 mM phenylmethylsulfonyl fluoride (PMSF) (Beyotime
   Biotechnology, ST506) for 20 min. After centrifugation at 13,000 g/min
   for 15 min at 4°C, the supernatant was obtained and subject to protein
   quantification with a BCA Protein Assay Kit (Beyotime Biotechnology,
   P0012) and then boiled for 10 min in 5× SDS PAGE sample loading buffer
   (Beyotime Biotechnology, P0015L). Equal protein samples were separated
   on 10% SDS-PAGE and transferred onto a polyvinylidene difluoride (PVDF)
   membrane. The PVDF membranes were first blocked in PBS containing 5%
   non-fat milk powder and 0.1% Tween 20 at 37°C for 1 h, which were then
   incubated with primary antibodies at 4°C overnight and then with the
   corresponding secondary antibodies conjugated to HRP at 37°C for 2 h.
   The immunolabeled protein complexes were visualized using an ECL Plus
   Kit (Beyotime Biotechnology, P0018), using the CanoScan LiDE 100
   scanner system (Canon).

   The gray-scale value of the specific bands was analyzed with ImageJ
   software for the semi-quantitation of protein expression.

Confocal immunofluorescence microscopy

   Cells were grown and treated in 35 mm Petri dishes (NEST, GBD-35-20)
   with a glass bottom. Then, the cells were washed with PBS and fixed
   with 4% paraformaldehyde (Sigma-Aldrich, P6148) for 30 min at room
   temperature and then permeated with 0.1% Triton X-100 (Sigma-Aldrich,
   T8787) for 10 min, followed by staining of nuclei with DAPI (Beyotime
   Biotechnology, C1002). Fluorescence signals were observed using a TCS
   SP2 confocal fluorescence microscope (Leica TCS SP8).

Statistical analysis

   The data are shown as the mean ± standard deviation of at least three
   independent experiments and were analyzed by one-way ANOVA or two-way
   ANOVA tests using the GraphPad Prism 9. P-values of less than 0.05 were
   considered to be statistically significant.

Supplementary Material

   Reviewer comments
   [222]reviewer-comments.pdf^ (569.9KB, pdf)

ACKNOWLEDGMENTS