TRIM72 restricts lyssavirus infection by inducing K48-linked ubiquitination and proteasome degradation of the matrix protein

The tripartite motif (TRIM) protein family is the largest subfamily of E3 ubiquitin ligases, playing a crucial role in the antiviral process. In this study, we found that TRIM72, a member of the TRIM protein family, was increased in neuronal cells and mouse brains following rabies lyssavirus (RABV) infection. Over-expression of TRIM72 significantly reduced the viral titer of RABV in neuronal cells and mitigated the pathogenicity of RABV in mice. Furthermore, we found that TRIM72 over-expression effectively prevents the assembly and/or release of RABV. In terms of the mechanism, TRIM72 promotes the K48-linked ubiquitination of RABV Matrix protein (M), leading to the degradation of M through the proteasome pathway. TRIM72 directly interacts with M and the interaction sites were identified and confirmed through TRIM72-M interaction model construction and mutation analysis. Further investigation revealed that the degradation of M induced by TRIM72 was attributed to TRIM72’s promotion of ubiquitination at site K195 in M. Importantly, the K195 site was found to be partially conserved among lyssavirus’s M proteins, and TRIM72 over-expression induced the degradation of these lyssavirus M proteins. In summary, our study has uncovered a TRIM family protein, TRIM72, that can restrict lyssavirus replication by degrading M, and we have identified a novel ubiquitination site (K195) in lyssavirus M.

Post-translational modifications (PTMs) of proteins, which involve to the addition or removal of covalent functional groups, play a crucial role in regulating protein structure, localization and activity [10].Common PTMs include phosphorylation, acetylation, ubiquitination, methylation, glycosylation, etc [11].Ubiquitination is an essential PTM process that covalently conjugates single or multiple ubiquitin molecules to one or more lysine residues of a substrate protein [12].This process is catalyzed by three types of enzymes: ubiquitin-activating enzymes (E1), ubiquitin-conjugating enzymes (E2) and ubiquitin-protein ligases (E3) [13].The ubiquitin-proteasome system plays a crucial role in degrading abnormal proteins and short-lived proteins within cells to maintain normal biochemical processes; it consists of ubiquitin, three types of ubiquitin-associated enzymes (E1, E2 and E3) and the proteasome [14].
More than 600 E3 ligases have been characterized and are classified into 3 different classes: homologous to E6-AP carboxyl terminus (HECT), really interesting new gene (RING) and RING-between RING (RBR) [15].RING proteins are the largest class of E3 ligases and among them the tripartite motif (TRIM) family of proteins was the largest subfamily of RINGs.TRIM proteins are involved in many biological processes, including transcriptional regulation, cell proliferation and differentiation, apoptosis, DNA damage repair, intracellular signal transduction and immune response, etc [16,17].TRIM proteins are characterized by the presence of an N-terminal tripartite or RBBC motif comprised of a RING domain, either one or two B-boxes (B1 and B2) and a coiled-coil (CC) domain, followed by a highly variable C-terminal domain [18,19].The C-terminal variable functional region of the TRIM protein plays a crucial role in substrate recognition thereby conferring functional specificity, the PRY-SPRY domain is the most common C-terminal domain among known TRIM proteins [20,21].
As a member of the TRIM protein family, TRIM72 (also known as MG53) is a RING-mediated E3 ubiquitin ligase that consists of five domains: the RING domain, B-box domain, coiled-coil domain, PRY domain and SPRY domain [22].The RING domain contains the E3 ubiquitin ligase activity responsible for mediating the ubiquitination of numerous proteins.For instance, TRIM72 directly interacts with Ras-related C3 botulinum toxin substrate 1

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release (RAC1) through its coiled-coil domain and suppresses RAC1 activity by catalyzing the Lys48 (K48)-linked polyubiquitination of RAC1 at Lys5 residue in HCC cells [23]; Additionally, TRIM72 induces ubiquitination of insulin receptor substrate 1 (IRS-1) with the assistant of an E2-conjugating enzyme UBE2H [24,25].The B-box domain, a zinc ion binding domain, plays a critical role in TRIM activity.A previous study has demonstrated that the B-box domain also regulates the interaction between TRIM72 and FAK8 [24].The conserved coiled-coil domain of TRIM72 is critical for TRIM72 homodimerization [26].Furthermore, the PRY-SPRY domain was shown to bind with Orai1 and regulate extracellular Ca 2+ entry [27].Consequently, TRIM72 is a multifunctional protein involved in various biochemical processes within cells.Our previous study revealed that TRIM25 can restrict RABV replication, indicating that proteins in the TRIM family play an important role in RABV infection [28].Recently, a study demonstrated that TRIM72 could protect mice from lethal influenza virus infection [29], indicating its potential antiviral function.However, the relationship between TRIM72 and the lyssaviruses remains unclear.
In this present study, we present the major discovery that RABV infection induces the upregulation of TRIM72 in neuronal cells and mouse brains.Over-expression of TRIM72 restricts the release of RABV in neuronal cells and reduces the pathogenicity of RABV in mice.We have determined that TRIM72 induces the degradation of RABV-M through the proteasome pathway.Our functional domain truncation analysis of TRIM72 indicates that the TRIM72-SPRY domain interacts with RABV-M.Furthermore, through the construction of a TRIM72-SPRY-M interaction model and mutation analysis, we have confirmed that TRIM72 ubiquitinates a specific site, K195, in M, thereby leading to its degradation.Significantly, we found that the K195 ubiquitination site is partially conserved among lyssavirus, and TRIM72 interacts with lyssavirus M proteins, ultimately resulting in the degradation of lyssavirus M proteins.Consequently, our study reveals a previously unrecognized TRIM72-ubiquitinationproteosome-mediated antiviral response and identified a novel ubiquitination site (K195) in lyssavirus M proteins.

RABV infection induces TRIM72 up-regulation in mouse brains and neuronal cells
To determine the relationship between RABV and TRIM72, we assessed the TRIM72 level in RABV-infected mouse brains and neuronal cells.For the in vivo experiments, 8-week-old C57 BL/6 mice were infected with 100 fluorescence focus units (FFU) RABV (CVS-B2c strain) via intranasal infection.As a negative control (Mock), the same volume of DMEM was treated.Brains were collected on the 12 th day post-infection.The mRNA level and protein level of TRIM72 were analyzed by qPCR and western blotting respectively.The mRNA level and protein level of TRIM72 were dramatically increased in RABV-infected mouse brains compared to the control group (Fig 1A and 1B).For the in vitro experiments, primary neuron cells were isolated and cultured, and then the neuron cells were infected with RABV at varying multiplicities of infection (MOI) for 36 hours (h) and the mRNA level and protein level of TRIM72 were analyzed by qPCR and western blotting.As shown in Fig 1C and 1D, the mRNA and protein levels of TRIM72 were gradually increased along with the RABV infection dose increased.Then mRNA level and protein level of TRIM72 were also analyzed in N2a cells (mouse neuroblastoma cell line) and SK-N-SH cells (human neuroblastoma cell line).As shown in Figs 1E, 1F, S1A and S1B, the mRNA level and protein level of TRIM72 or human TRIM72 (hTRIM72) gradually increased along with the RABV infection dose increased in N2a cells or SK-N-SH

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release

TRIM72 suppresses RABV replication in neuronal cells
To further investigate the relationship between TRIM72 and RABV, over-expression plasmids of TRIM72-flag and TRIM72 knockout N2a cell line (Trim72 -/-) were constructed.We then transfected the TRIM72 over-expression vector into N2a cells for 48 h, and the TRIM72-flag overexpression level was analyzed by western blotting and indirect immunofluorescence.In comparison to the empty vector group, the TRIM72 level was dramatically increased in the TRIM72 over-expression group (Figs 1G and S1C).To confirm the deficiency of TRIM72 in TRIM72 -/-N2a cells, the TRIM72 expression level was also analyzed by comparing the expression of TRIM72 between WT and TRIM72 -/-N2a cells by western blotting.No TRIM72 protein expression was detected in TRIM72 -/-N2a cells (Fig 1H ), indicating that TRIM72 is well knocked out in the constructed TRIM72 -/-N2a cells.In addition, no effect on cell viability was observed after over-expressing or knocking out TRIM72 (S1D and S1E Fig) .Next, an empty vector or TRIM72 over-expression vector was transfected into N2a cells for 12 h and then infected the cells with RABV (MOI = 0.01).The supernatants of RABV-infected cells were collected at 24 h and 48 h post-infection, then viral titers were analyzed.From the results we can see RABV titer was dramatically decreased post TRIM72 over-expression no matter at 24 h or 48 h post-infection (Fig 1I).Moreover, to find out the proper expression level of exogenous TRIM72 that matches the endogenous expression level of TRIM72 during viral infection, a dose-dependent TRIM72 expression experiment was carried out by transfecting different concentrations of TRIM72 expressing vectors (0.25 μg, 0.5 μg, 0.75 μg, and 1.0 μg) into N2a cell, and the expression levels of TRIM72 was determined to compare with the expression level of TRIM72 during RABV infection.It was found that the TRIM72 expression level in N2a cells transfected with 0.5 μg of TRIM72 expressing vector presents a comparable level of TRIM72 in RABV infected N2a cells at 48 h post-infection (S1F Fig) .Meanwhile, a dose-dependent viral inhibition study was also carried out by transfecting different concentrations of TRIM72 into N2a cells, and the viral titers were then determined to assess the inhibitory effect on RABV infection by different expression levels of TRIM72.As shown in S1G Fig, a significant decrease of viral titer was observed in N2a cells transfected with 0.5 μg of TRIM72 expressing vector, which is determined as a comparable expression level of TRIM72 with that during RABV infection.To analyze the effect of Trim72 deficiency on RABV replication in N2a cells, WT N2a cells and Trim72 -/-N2a cells were infected with RABV (MOI = 0.01) respectively, and the cell supernatants were collected at 24 h and 48 h post-infection and viral titers were analyzed.From the results, we can see that RABV titer dramatically increased in the Trim72 -/-group compared with the WT group both at 24 h and 48 h post-infection (Fig 1J).Furthermore, the impact of TRIM72 deficiency on RABV was abolished by over-expression of TRIM72 (Fig 1K).Moreover, we found that human TRIM72 (hTRIM72) over-expression dramatically TRIM72 antibody.(I) Empty vector or TRIM72-flag over-expression vectors were transfected into N2a cells respectively for 12 h, then infected with RABV (MOI = 0.01) for the indicated time, and the viral titers in the supernatant were analyzed.(J) WT and Trim72 -/-N2a cells were infected with RABV (MOI = 0.01) for the indicated time respectively, and the viral titers in the supernatant were analyzed.(K) WT and Trim72 -/-N2a cells were transfected with empty vector or TRIM72-flag over-expression vectors for 12 h, then infected with RABV (MOI = 0.01) for 48 h and the viral titers in the supernatant were analyzed.(L) Trim72 and Trim72(-) were inserted into the genome of a recombinant RABV (rRABV), named rRABV-TRIM72 and rRABV-TRIM72(-) respectively.(M) N2a cells were infected with different types of rRABVs (MOI = 3) for 36 h, then the TRIM72 level was analyzed by western blotting with an anti-TRIM72 antibody.(N) N2a cells were infected with different types of rRABVs (MOI = 0.01) and their growth kinetics were compared.Statistical analysis of grouped comparisons was carried out by student's t-test (*P < 0.05; **P<0.01;***P<0.001;****P<0.0001).The bar graph represents means ± SD, n = 3.Western blot data are representative of at least two independent experiments.https://doi.org/10.1371/journal.ppat.1011718.g001

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release reduced the RABV titer in SK-N-SH cells (S1H and S1I Fig) .To further analyze the function of TRIM72 in restricting RABV, the recombinant RABVs (rRABVs) were constructed.The recombinant viruses were derived from the CVS-B2c strain, including unaltered rRABV, rRABV harboring TRIM72 coding sequence (rRABV-Trim72) and rRABV harboring the TRIM72 sequence which ATG mutation to CTG (rRABV-Trim72(-)) (Fig 1L).N2a cells were infected with those rRABVs (MOI = 3) for 36 h, then the TRIM72 expression level was analyzed by western blotting.As expected, the TRIM72 protein level was extremely higher in rRABV-TRIM72 infected N2a cells than in rRABV or rRABV-TRIM72(-) infected cells (Fig 1M).Subsequently, virus growth kinetics experiments were performed in N2a cells, and N2a cells were infected with the rRABVs (MOI = 0.01) separately, the supernatants were collected at the indicated time points post rRABVs infection for analysis of virus titers.The virus titer was significantly lower in the rRABV-TRIM72 infected cells than in rRABV-and rRABV--TRIM72(-)-infected cells (Fig 1N).Above all, over-expression of TRIM72 reduced the viral titer of RABV in neuronal cells.

TRIM72 reduces RABV pathogenicity in vivo
To investigate the role of TRIM72 in RABV infection in vivo, the pathogenicity of rRABV, rRABV-Trim72 and rRABV-Trim72(-) in the C57 BL/6 mouse model was compared.Mice were infected intranasally with rRABV, rRABV-Trim72, or rRABV-Trim72(-) at a dose of 100 FFU/mouse.Then the body weight, clinical score and survival were monitored.The body weight of the mice infected with rRABV or rRABV-Trim72(-) decreased dramatically from day 7 to day 13 post-infection (p.i.), whereas the body weight of those mice infected with rRABV-Trim72 showed a slight decrease from day 13 to day 17 p.i. (Fig 2A).The rabies clinical symptoms of rRABV-or rRABV-Trim72(-) -infected mice were appeared from day 6 to day 13, whereas the mice infected with rRABV-TRIM72 display mild symptoms from day 11 to day 17 (Fig 2B).Finally, 80% of the mice infected with rRABV-Trim72 survived, and only 10% and 20% survival ratio for rRABV-and rRABV-Trim72(-) -infected mice, respectively (Fig 2C).
To analyze the viral loads in the brains of mice infected with rRABV, rRABV-Trim72 and rRABV-Trim72(-), the RABV viral N mRNA levels were analyzed in different brain parts followed by intranasal viral infection (100 FFU/mouse).On day 6, day 9 and day12 p.i., the brains infected with different viruses were collected and divided into 5 parts (Cerebrum, Thalamus, Midbrain, Cerebellum and Brainstem), then the N mRNA levels in different brain parts were analyzed by qPCR.Dramatically reduced RABV N mRNA levels in rRABV-Trim72-infected different mice brain parts were presented compared with rRABV-or rRABV-Trim72 (-)-infected mouse brains at day 9 and day 12 p.i. (Fig 2D -2H).Then the immunohistochemistry assay was performed by using an anti-RABV-P antibody to analyze the RABV protein level in different brain parts post virus infection for 12 days.Almost no RABV-P antigen was observed in rRABV-Trim72 infected brain parts, whereas comparable levels of RABV-P antigen have been observed in rRABV-or rRABV-Trim72(-)-infected mice brain parts (Fig 2I).Above all, over-expression of TRIM72 dramatically reduced the pathogenicity of RABV in mice.

TRIM72 restricts RABV assembly and/or release
Our results demonstrate that TRIM72 inhibits RABV in vitro and in vivo, while the mechanism remains unclear.TRIM proteins play essential roles in various biological processes, especially in innate immune response [30].Thus the effect of TRIM72 on type I interferon (IFN) system was initially analyzed.As shown in S2A-S2E Fig, the mRNA levels of IFN-β and some

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release important ISGs such as ISG15 or IFIT1/2/3 were decreased post TRIM72-flag over-expression during RABV infection.Reversely, the mRNA levels of IFN-β, ISG15 or IFIT1/2/3 were increased in Trim72 -/-N2a cells during RABV infection (S2F-S2J Fig) .Furthermore, the restriction effect of TRIM72 on RABV was analyzed in Vero cell which is defective in the interferon synthesis system [31,32].As shown in S2K Fig, TRIM72-flag over-expression in Vero cells reduces the viral titer.Moreover, the virus growth kinetics experiments were performed in Vero cells, and Vero cells were infected with the rRABVs (MOI = 0.01) separately, the supernatants were collected at the indicated time points post rRABVs infection for analysis of virus titers.The virus titer was significantly lower in the rRABV-TRIM72 infected cells than in rRABV-and rRABV-TRIM72(-) infected cells (S2L Fig).Above all, these results indicate that TRIM72 plays a negative control on the type I IFN system and the IFN system is not the major mechanism for the restriction effect of TRIM72 on the viral titer of RABV.
Therefore, the effects of TRIM72 over-expression on the different stages of the RABV life cycle were analyzed.First, the TRIM72-flag over-expression vector or an empty vector was individually transfected into N2a cells for 36 h and then the cells were infected with RABV (MOI = 0.1) for 1 h at 4˚C, the cells were collected and the viral genomic RNA (vRNA) level was analyzed by qPCR to evaluate the effect of TRIM72 on RABV attachment.As shown in Fig 3A, there is no obvious difference in viral genomic RNA levels between the control group and the TRIM72-flag over-expression group, indicating that TRIM72 did not affect RABV attachment.Next, the effect of TRIM72 over-expression on RABV entry was analyzed, TRIM72-flag over-expression vector or empty vector was individually transfected into N2a cells for 36 h and then infected with RABV (MOI = 0.1) for 2 h at 37˚C, the cells were collected and the vRNA levels were analyzed by qPCR.There is no obvious difference in vRNA levels between the control group and the TRIM72 over-expression group (Fig 3B ), indicating that TRIM72 did not affect RABV entry.Then TRIM72-flag over-expression vector or empty vector was individually transfected into N2a cells for 12 h and then RABV was infected (MOI = 0.01), the cells were collected at 24 h and 48 h post-infection, and the RABV N mRNA level and vRNA level were analyzed by qPCR.As shown in Fig 3C and 3D, there is almost no difference in RABV N mRNA level and vRNA level between the TRIM72 over-expression group and the empty vector group.These results indicate that TRIM72 over-expression does not affect RABV transcription and replication.Then virus-spreading assay was performed to analyze the impact of TRIM72 on RABV spreading between cells.N2a cells were infected with rRABV, rRABV--Trim72, or rRABV-Trim72(-) individually (MOI = 0.001), then the cell culture medium containing 1% agar was added to cover the N2a cells for 48 h.Fluorescence in isothiocyanate (FITC)-conjugated anti-RABV-P protein antibody (FITC-P) was stained post removed from the cell culture medium and the fluorescence plaque was analyzed.There is almost no difference in the size of fluorescence plaque and the number of infected cells in these N2a cells (Fig 3E and 3F), indicating that TRIM72 does not affect RABV spreading between cells.
Our results clearly show that TRIM72 over-expression dramatically reduced the RABV titer in N2a cells (Figs 1I and S1I).However, TRIM72 over-expression almost does not affect RABV attachment, entry, transcription, replication and spreading in N2a cells.Therefore, we hypothesized that TRIM72 might affect RABV assembly and/or release.To investigate the hypothesis, a virus-like particle (VLP) system was constructed as previously reported to analyze the effect of TRIM72 over-expression on RABV assembly and/or release [33,34].TRIM72-myc overexpression vector or empty vector together with RABV-M-HA and RABV glycoprotein (G)flag overexpression vectors were co-transfected into 293T cells for 48 h and the VLP production in the supernatants was analyzed using western blotting.As shown in Fig 3G, the protein levels of M and G in the supernatants were dramatically decreased post-TRIM72 over-expression compared to the control group, indicating that TRIM72 restricts VLP production.As our

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release speculation, these results confirm our hypothesis that TRIM72 over-expression restricts RABV assembly and/or release.

TRIM72 induces proteasomal degradation of RABV-M by promoting its K48-linked ubiquitination
Our results revealed that TRIM72 restricts RABV assembly and/or release.However, an interesting observation was the significant decrease in M protein levels following TRIM72 overexpression (Fig 3G).Given that TRIM72 is an E3 ubiquitin ligase, we speculated that its inhibitory effect on RABV may be responsible for its E3 ubiquitin ligase function, leading to the degradation of the viral proteins.To explore this hypothesis, over-expression vectors of RABV viral proteins (N, P, M, G), together with TRIM72-myc expression vector or an empty vector, were individually co-transfected into N2a cells for 48 h, the cells were collected and the viral protein levels were analyzed by western blotting.As shown in Fig 4A, the protein levels of RABV-N, -P, or -G-flag almost have no changes following TRIM72-myc over-expression.In contrast, the protein level of RABV-M-HA was dramatically decreased in the TRIM72-myc over-expression group compared to the empty vector group.Furthermore, RABV-M level was  We then speculated that TRIM72-induced degradation of M via the proteasome pathway was due to its E3 ubiquitin ligase function which promotes M ubiquitination.The RING domain is essential for TRIM72 to promote the ubiquitination of the target proteins [25].Thus, TRIM72, a RING domain deficient TRIM72 (TRIM72-ΔRING) together with M were co-overexpressed in N2a cells for 48 h and the M level was analyzed by western blotting.As shown in Fig 4F, WT TRIM72 over-expression induces M degradation, whereas TRIM72ΔR-ING over-expression failed to induce M degradation.The phosphorylation of the S255 site in TRIM72 has been demonstrated as a prerequisite for TRIM72 E3 ligase activity [35].Then we mutated the Serine at 255 site of TRIM72 to Alanine to analyze whether the effect of TRIM72 on RABV M degradation was dependent on its E3 ligase activity.As shown in S3B Fig, the degradation of RABV-M induced by TRIM72 was abolished with the introduction of the S255A mutation.Moreover, the restriction effect of TRIM72 on RABV was also abolished with the S255A mutation (S3C Fig) .These results indicate that TRIM72 induces M degradation depending on its RING domain and E3 ligase activity.We then examined the effect of TRIM72 on the ubiquitination of M. The ubiquitination level of M was dramatically increased following TRIM72 over-expression in the presence of Mg132 (Fig 4G ), indicating that TRIM72 promotes the ubiquitination of M. We then tested the effect of TRIM72 on the ubiquitination level of RABV-M during RABV infection.N2a cells were transfected with empty vectors or TRIM72-flag expressing vectors for 12 h and then infected with RABV (MOI = 1) for 36 h.The endogenous Co-IP was then performed with an anti-M antibody.As shown in Fig 4H, the ubiquitination level of RABV-M was higher in the TRIM72 over-expression group compared with the empty vector group (Fig 4H).K48-and K63-linked polyubiquitin chains are the two major types of ubiquitin linkages [36], then we further characterized which type of ubiquitination of M was modified by TRIM72.TRIM72 over-expression plasmids and M transfected into N2a cells for 48 h.Then the protein levels of TRIM72-flag and M-HA were analyzed by western blotting.(E) pCA-M-HA together with pCAGGS or different concentrations of pCA-Trim72-flag were co-transfected into N2a cells.Then Mg132 (10 μM) was applied and the protein levels of TRIM72-flag and M-HA were analyzed by western blotting.(F) pCA-M-HA together with pCAGGS, pCA-Trim72-flag, or Trim72-ΔRING-flag were cotransfected into N2a cells respectively for 48 h, then M-HA level was analyzed by western blotting.(G) pCAGGS or pCA-Trim72-myc together with pCA-Mflag and UB-HA over-expression vectors were co-transfected into N2a cells respectively.Then treated with Mg132 and Co-IP assays were performed with antiflag antibody.The protein levels of TRIM72-myc, M-flag, and UB-HA were analyzed by western blotting.(H) pCAGGS or pCA-Trim72-flag were transfected into N2a cells respectively.Then infected with RABV (MOI = 1) for 36 h and treated with Mg132 and Co-IP assays were performed with anti-M antibody.The protein levels of TRIM72-flag, M, and UB were analyzed by western blotting.(I) pCAGGS or pCA-Trim72-myc together with pCA-M-flag and UB-K48-HA over-expression vectors were co-transfected into N2a cells respectively.Then treated with Mg132 and Co-IP assays were performed with anti-flag antibody post transfection for 48 h.Then the protein levels of TRIM72-myc, M-flag, and UB-K48-HA were analyzed by western blotting.(J) pCAGGS or pCA-Trim72-myc together with pCA-M-flag and UB-K63-HA over-expression vectors were co-transfected into N2a cells respectively.Then treated with Mg132 and Co-IP assays were performed with anti-flag antibody.The protein levels of TRIM72-myc, M-flag, and UB-K63-HA were analyzed by western blotting.Western blot data are representative of at least two independent experiments.https://doi.org/10.1371/journal.ppat.1011718.g004

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release over-expression plasmids together with either K48-or K63-ubiquitin over-expression plasmids were co-transfected into N2a cells for 48 h respectively, then ubiquitination assays were performed.As shown in Fig 4I and 4J, TRIM72 specifically promotes the K48-linked ubiquitination of M, while almost does not affect the K63-linked ubiquitination of M. Overall, TRIM72 promotes K48-linked ubiquitination of M, consequently facilitating M degradation via the proteasome pathway.

The SPRY domain of TRIM72 directly interacts with RABV-M
Our above results demonstrate that TRIM72 promotes the K48-linked ubiquitination of RABV-M, thus we speculated that TRIM72 might interact with RABV-M directly.Consequently, Co-Immunoprecipitation (Co-IP) assays were performed to investigate this interaction, as shown in Figs 5A-5C, S4A and S4B, both murine TRIM72 and hTRIM72 were found to interact directly with M no matter under over-expression condition or RABV infection condition.For further analysis of the interaction between TRIM72 and RABV-M, different TRIM72 truncation over-expression vectors were designed and constructed based on its functional domain (Fig 5D).These vectors together with M over-expression vectors were co-overexpressed in N2a cells for 48  We then attempted to construct an M-TRIM72-SPRY interaction model to analyze the specific interaction sites.Since there was no available structure model for RABV-M, thus we built a RABV-M model based on the Lagos bat virus M protein structure with SWISS-MODEL online software.The RABV-M model is shown in Fig 5H .A previous study has reported the hTRIM72 crystal structure, allowing us to employ it as the model for building the M-TRIM72-SPRY interaction model using GalaxyWEB online software.The M-TRIM72-SPRY interaction model is shown in Fig 5I , and the potential interaction sites between M and TRIM72-SPRY were analyzed and presented.We found that the residues R404, D421 and VP436-437 of hTRIM72-SPRY could interact with the residues K60, PN64-65, M110, DT147-148 and Q154 of M respectively.The potential interaction residues were then validated by mutation and Co-IP assays.As shown in Fig 5J, the protein level of SPRY that was pulled down by mutated M (PN64-65AA, M110A, DT147-148AA and Q154A) was dramatically decreased compared to that of WT M. Since the amino acid sequence homology between hTRIM72-SPRY and murine TRIM72-SPRY reached 84.3% (Fig 5K).Therefore, the corresponding interaction residues of the murine TRIM72-SPRY domain were mutated and Co-IP experiments were performed to determine the changes in binding between the mutated murine TRIM72-SPRY and M. As shown in Fig 5L, the M protein level pulled down by mutated TRIM72-SPRY (R404A, D421A and VP436-437AA) dramatically reduced compared to that of WT TRIM72-SPRY.In conclusion, these results strongly support the reliability of the key interaction sites between M and TRIM72-SPRY as analyzed by the predicted M-TRIM72-SPRY interaction model.Specifically, residues PN64-65, M110, DT147-148 and Q154 on M, as well as R404, D421, and VP436-437 on TRIM72-SPRY, have been identified as critical for the interaction between TRIM72-SPRY and M.

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release Furthermore, we sought to analyze whether these interaction site mutations in TRIM72-SPRY disrupted the function of TRIM72 in M degradation, then the over-expression vectors of full-length TRIM72 mutations (R404A, D421A and VP436-437AA) were constructed and together with M over-expression vector were co-transfected into N2a cells for 48 h.Then the M levels were analyzed by western blotting.Interestingly, the protein level of M was comparable in the TRIM72 mutated group compared to the empty vector group, while the M level decreased significantly following WT TRIM72 over-expression (Fig 5M).Moreover, the antiviral function of TRIM72 mutations was also analyzed.The over-expression vectors of WT full-length TRIM72 or mutations were transfected into N2a cells for 12 h and then infected with RABV (MOI = 0.01) for 48h, the virus titers in the cell culture supernatant were measured.As shown in Fig 5N, the antiviral function of TRIM72 was impaired following R404, D421 and VP436-437 mutation in TRIM72 compared to the WT TRIM72 group.

TRIM72 promotes the ubiquitination of RABV-M at the K195 site
To identify the specific lysine site in M that is ubiquitinated by TRIM72, the lysine in M was mutated respectively (Fig 6A).The over-expression vectors of M which lysine (K) mutated to alanine (A) were constructed and together with TRIM72 were co-overexpressed in N2a cells for 48 h respectively, then western blots were performed and the protein level of mutated M was analyzed.As shown in Fig 6B, only the M-K195A mutant abolished the degradation of M which was induced by TRIM72, indicating that the K195 is a potential lysine residue targeted by TRIM72 for M ubiquitination.Subsequently, the stability of WT M and M-K195A were compared when individually over-expressed in N2a cells for 48 h, the M level was analyzed by western blot.As shown in Fig 6C , The M-K195A protein level was higher than that of WT M, indicating that M-K195A was much more stable than WT M. Then we examined the ubiquitination level of WT M and M-K195A which was induced by TRIM72.Empty vector or TRIM72 over-expression vector together with ubiquitin and WT M or M-K195A were cooverexpressed in N2a cells for 48 h and then the ubiquitination level of WT M and M-K195A was analyzed.As shown in Fig 6D , WT M presents a decreased ubiquitination level in TRIM72 over-expression group compared with the empty vector group, and M-K195A presents a comparable ubiquitination level between the vector group and TRIM72 over-expression group.Moreover, the ubiquitination level of M-K195A was lower compared with WT M.These Then Co-IP assays were performed with anti-TRIM72 antibody.The protein levels of TRIM72 and M were analyzed by western blotting.(D) TRIM72 truncations were designed and constructed based on its functional domain.(E) The over-expression vectors of TRIM72-flag truncations together with pCA-M-HA were co-transfected into N2a cells respectively for 48 h.Then Co-IP assays were performed with antiflag antibody.The protein levels of TRIM72 truncations and M-HA were analyzed by western blotting.(F) RABV-M truncations were designed and constructed based on its secondary structure.(G) The over-expression vectors of M-HA truncations together with pCA-Trim72-SPRY-flag were co-transfected into N2a cells for 48 h.Then Co-IP assays were performed with anti-HA antibody.The protein levels of M truncations and SPRY-flag were analyzed by western blotting.(H) A structural model of RABV-M was built using SWISS-MODEL online software (https://swissmodel.expasy.org/interactive)based on the crystal structure of Lagos bat virus M protein (PDB code: 2W2S).(I) An interaction model of the human TRIM72-SPRY domain and RABV-M was built with GalaxyWEB online software (https://galaxy.seoklab.org/) based on hTRIM72 structure (PDB code: 6NPY), the potential interaction sites were labeled.(J) Over-expression vectors of RABV-M-HA mutations together with TRIM72-SPRY-flag over-expression vector were co-transfected into N2a cells respectively for 48 h.Then Co-IP assays were performed with anti-HA antibody.The protein levels were analyzed by western blotting.(K) The protein sequences of SPRY domain between hTRIM72 and mouse TRIM72 were compared and analyzed with ESPript 3.0 online software (https://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi).(L) Over-expression vectors of TRIM72-SPRY-flag mutations together with M-HA over-expression vectors were co-transfected into N2a cells respectively for 48 h.Then Co-IP assays were performed with anti-flag antibody and protein levels were analyzed by western blotting.(M) Over-expression vectors of TRIM72-flag mutations together with M-HA over-expression vector were cotransfected into N2a cells respectively for 48 h.The protein levels were analyzed by western blotting.

PLOS PATHOGENS
TRIM72 lyssavirus assembly and/or release results demonstrated that the K195A mutation abolished the enhanced ubiquitination level induced by TRIM72.These results strongly indicate that the K195 site is a critical ubiquitination target for M, mediated by TRIM72.
To further investigate the impact of the M-K195A mutation on the effect of TRIM72 on RABV assembly and release, the VLP production assay was performed.TRIM72 or empty vector together with M-K195A and RABV G were co-overexpressed in 293T cells for 48 h and the VLP production in the supernatants was analyzed by western blotting.As shown in Fig 6E, the protein levels of M-K195A and G in the supernatants were comparable between the TRIM72 over-expression group and the control group, indicating that TRIM72 lost its ability to restrict the VLP production following M-K195 site mutation.Moreover, A M-K195A mutant RABV (rRABV-M-K195A) was constructed (Fig 6F ), and virus growth kinetics experiments were performed in N2a cells.Briefly, N2a cells were infected with rRABV or rRABV-M-K195A (MOI = 0.01), and the supernatants were collected at the indicated time points post rRABVs infection for virus titration.The virus titers were significantly higher in the rRABV-M-K195Ainfected cells than in rRABV-infected cells at all tested time points (Fig 6G).Then the restriction effect of TRIM72 on rRABV-M-K195A was analyzed.N2a cells were transfected with either a vector or TRIM72 over-expression vector for 12 h and then infected with rRABV or rRABV-M-K195A (MOI = 0.01).The supernatants were collected and viral titers were analyzed at 48 h post-infection.As shown in Fig 6H , TRIM72 over-expression reduced the viral titer of rRABV, while there were almost no obvious changes in viral titers between the vector group and the TRIM72 over-expression group during rRABV-M-K195A infection.These results indicate that TRIM72 restricts RABV assembly and release by promoting the ubiquitination of M at the 195 site, thereby promoting the degradation of the M protein.

TRIM72 targeting the K195 site is partially conserved among lyssavirus M
RABV is a member of lyssaviruses, the M proteins display a high homology among lyssaviruses (S5 Fig) .Therefore, we speculated that TRIM72 might also induce the degradation of other lyssavirus M proteins.Our results demonstrate that TRIM72 promotes CVS-M-K195 ubiquitination thereby promoting the degradation of CVS-M protein.Then the K195 site was analyzed among the M proteins of those lyssaviruses.As shown in S5 Fig, the K195 site was partially conserved among those lyssavirus M proteins, such as dog-derived RABV M (named DRV-M), ABLV-M (At the 205 site in ABLV-M), DUVV-M and EBLV-M, while the 195 site was the E instead of K in LBV-M or MOKV-M.Firstly, we tested the effect of TRIM72 on DRV-M, the DRV-M together with TRIM72 or empty vector were co-overexpressed in N2a cells for 48 h, and then western-blot was performed to analyze the DRV-M protein level.As shown in Fig 7A , the DRV-M protein level decreased significantly following TRIM72 overexpression, indicating that TRIM72 induce the degradation of DRV-M.The other M proteins containing the K195 site of these lyssaviruses were also tested including ABLV-M, DUVV-M and EBLV-M.As shown in Fig 7B -7D, similar results were observed, where the protein levels of ABLV-M, DUVV-M and EBLV-M were dramatically reduced following TRIM72 overexpression, indicating that TRIM72 induced the degradation of those lyssavirus M proteins.Importantly, treatment with Mg132, a proteasome inhibitor, could reverse the degradation of lyssavirus M proteins induced by TRIM72 over-expression (Fig 7E -7H).Moreover, the results of Co-IP assays demonstrated that DRV-M, ABLV-M, DUVV-M and EBLV-M could also interact with TRIM72 (Fig 7E -7H).However, when we tested the effect of TRIM72 on LBV-M analyzed.Statistical analysis of grouped comparisons was carried out by student's t-test (*P < 0.05; **P<0.01;***P<0.001;****P<0.0001).The bar graph represents means ± SD, n = 3.Western blot data are representative of at least two independent experiments.https://doi.org/10.1371/journal.ppat.1011718.g006

PLOS PATHOGENS
TRIM72 restricts lyssavirus assembly and/or release and MOKV-M, we found that although TRIM72 could interact with LBV-M or MOKV-M, TRIM72 did not induce the degradation of LBV-M or MOKV-M (S6A and S6B Fig) .Then the K195 (or at the 205 site in ABLV-M) sites of these lyssaviruses' M proteins were mutated to Alanine.The mutated M proteins were co-overexpressed with TRIM72 or empty vector in N2a cells for 48 h, and the levels of mutated M proteins were analyzed by western blot.As shown in Fig 7I -7L, the levels of mutated M proteins were comparable between the TRIM72 over-expression group and the vector group, indicating that the mutation of the K195 (At the 205 site in ABLV-M) in lyssavirus M proteins impedes the degradation induced by TRIM72.When we mutate the 195 site in LBV-M or MOKV-M from E to K, TRIM72 can induce the degradation of LBV-M-E195K or MOKV-M-E195K (S6C and S6D Fig) .Overall,

PLOS PATHOGENS
TRIM72 lyssavirus assembly and/or release the K195 (or at 205 site in ABLV-M) ubiquitination sites were partially conserved among lyssavirus M proteins and were essential for the degradation induced by TRIM72.

Discussion
The TRIM protein family is a large family consisting of more than 80 members, and TRIM proteins play important roles in innate and adaptive immune responses to defend against viral invasions [30].Moreover, TRIM proteins can directly target viral proteins, leading to the degradation or functional inhibition of these viral proteins, thereby disturbing the viral lifecycle [30].In this study, we have provided evidence demonstrating that TRIM72 restricts the release of lyssavirus by inducing the degradation of lyssavirus M protein in neuronal cells.First, we found that TRIM72 was up-regulated following lyssavirus infection in neuronal cells and mouse brains.Subsequently, we identified a direct interaction between TRIM72 and lyssavirus M, leading to the promotion of K48-linked ubiquitination of lyssavirus M at the K195 site, thereby facilitating lyssavirus M degradation via proteasome pathway, effectively restricting the assembly and/or release of lyssavirus (Fig 8).Further study has revealed that the K195 site within M protein is partially conserved among lyssavirus M proteins.Importantly, these M proteins of lyssavirus that contain the conserved K195 site can be effectively degraded by TRIM72.This study is the first report demonstrating that lyssavirus M proteins undergo ubiquitination and degradation through the TRIM72-dependent proteasomal pathway.
TRIM72 plays an important role in membrane damage repair [37], and serves a protective role in many tissues, including neurons, heart, liver, kidney and muscle [38][39][40][41][42][43].A previous study has indicated that TRIM72 attenuated LPS-induced neurotoxicity and neuroinflammation by inhibiting TLR4/NF-κB pathway both in vitro and in vivo [44].A recent study has revealed that TRIM72 reduces muscle inflammation by promoting the inactivation of NLRP3 inflammasome [45].Additionally, several studies have found a connection between TRIM72

PLOS PATHOGENS
TRIM72 lyssavirus assembly and/or release and the influenza virus, demonstrating that TRIM72 protects mice from lethal influenza virus infection by suppressing interferon-β and inflammation [29,46].However.Until our study, there had been no reports of a direct interaction between TRIM72 and viral proteins.Our study here reveals a direct interaction between TRIM72 and lyssavirus M protein.Moreover, the interaction between TRIM72 and lyssavirus M promotes the degradation of lyssavirus M via a K48-linked ubiquitination modification in M at the K195 site.Recent studies have shown that TRIM72 is a potential E3-ligase enzyme [24,25], and our findings in this study provide significant evidence to support this conclusion, confirming TRIM72 as an E3-ligase enzyme.
Ubiquitination is an important post-translational modification (PTM) that plays a key role in a wide range of cellular biological processes, including signal transduction [47], immune response [48], cancer [49], metabolism [50], etc.With a deeper understanding of ubiquitination, multiple forms of ubiquitination have been revealed, including K6-, K11-, K27-, K29-, K33-, K48-and K63-linked polyubiquitinations [51,52].Among the numerous polyubiquitination types, K48-linked polyubiquitination and K63-linked polyubiquitination have been the most extensively studied, multiple studies have proved that K48-linked polyubiquitination induces the degradation of target proteins and K63-linked polyubiquitination regulates the activation of target proteins [36,51,53].K48-linked polyubiquitinations, which mediate proteasomal degradation, are the most abundant connection between ubiquitin proteins in cells [54].The K48-linked polyubiquitination, responsible for the degradation of target proteins, plays essential roles in numerous biological processes, whether involving host or viral proteins.TRIM21, an interferon-inducible E3 ligase, has been shown to induce K48-linked ubiquitination and subsequent degradation of DDX41, thereby negatively regulating the innate immune response [55].A most recent report has demonstrated that TRIM21 also restricts influenza A virus replication via a K48-linked degradation of the influenza A virus M1 protein [56].These reports indicate that TRIM proteins could modulate the degradation of both host and viral proteins.TRIM72 has previously been reported to induce the K48-linked ubiquitination and degradation of FAK while regulating skeletal myogenesis [24].However, there had been no prior reports of an association between TRIM72 and viral protein.In this study, we have demonstrated that TRIM72 directly interacts with lyssavirus-M and promotes K48-linked polyubiquitination at the lyssavirus-M K195 site, thereby promoting the degradation of lyssavirus M via the proteasome.
Lyssavirus M is the most abundant and smallest protein within the lyssavirus virion [57].Lyssavirus M plays an essential role in the lyssavirus life cycle, with its primary functions encompassing viral assembly/budding and the regulation of the balance between transcription and replication [7,8].As a multiple-functional protein, lyssavirus M is involved in numerous biological processes during lyssavirus infection.M displays different functions via direct or indirect interaction with different host proteins and several important functions have been identified including defense host immune response, induce apoptosis and regulate lyssavirus replication [58].NF-κB pathway is essential for defense against viral infection, and antiviral cytokines are produced once NF-κB is activated during virus infection.With this context, RelAp43, a member of the NF-κB pathway, has been found to participate in the innate immune response against lyssavirus infection [59].Recent studies have revealed that lyssavirus M disturbs the NF-κB pathway by interacting with RelAp43 [9,60].Additionally, lyssavirus M was also found to interact with Jak1 and cooperate with lyssavirus P to modulate the Jak-Stat pathway, thereby restricting the expression of interferon-stimulated genes (ISGs) [61].Lyssavirus induces apoptosis in neuronal cells during infection [62], and lyssavirus M contributes to this process via interacting with Cco1, ultimately decreasing the viability of Cco1, thereby disrupting the mitochondrial morphology and inducing apoptosis [63].Furthermore, lyssavirus M facilitates its replication via interacting with host proteins, a recent study has shown that M interacts with ATP6V1A to promote lyssavirus uncoating [34].
TRIM proteins display an essential role in defending against virus invasion, while viruses also evolved functions to hijack the host ubiquitin system to facilitate their life cycle at different stages, including virus entry, virus replication, virus assembly and budding [64].For example, TRIM7 induces K63-linked polyubiquitination on the E protein of Zika virus (ZIKV), the ubiquitination enhances ZIKV entry into cells via direct interaction with TIM1 a host cell receptor [65].TRIM6 ubiquitinates the K309 site in the VP35 protein of Ebola virus (EBOV) and the ubiquitination facilitates the polymerase cofactor activity thereby enhancing the replication of EBOV [66].Previous studies have shown that the ubiquitination of lyssavirus M which is regulated by the host is essential for lyssavirus budding [67,68].However, there are fewer reports on E3 ligases involved in lyssavirus M ubiquitination except for NEDD4 which is a membrane-localized ubiquitin ligase containing multiple WW domains that interact with the PPxY domain in the M protein of RABV [68].Although the interaction sites between M and TRIM72-SPRY were conserved among lyssavirus M proteins (S5 Fig) , it is worth noting that the degradation of lyssavirus M induced by TRIM72 is dependent on the ubiquitination at the K195 site, thus TRIM72 cannot induce the degradation of some lyssavirus M proteins due to the deficient of lysine at 195 site, such as LBV-M and MOKV-M.According to our results, there is a higher ubiquitination level in lyssavirus M following mutation at the M K195 site (Fig 6D ), indicating that there exist other ubiquitination sites in lyssavirus M.Although a previous study has identified a ubiquitination site at K60 of RABV M protein [69], additional ubiquitination sites may exist in lyssavirus M, necessitating further investigation.
In summary, our study has revealed a previously unknown role of TRIM72 during lyssavirus infection.We found that lyssavirus infection triggers an increase in TRIM72 levels in neuronal cells and mouse brains.Furthermore, we have identified a novel function of TRIM72, wherein it restricts lyssavirus assembly and/or release via directly interacting with lyssavirus M protein and facilitating the ubiquitination of lyssavirus's M proteins at the K195 site, thereby promoting the degradation of lyssavirus M. The discovery in this study is significant because it sheds light on the direct interaction between TRIM72 and viral proteins, it also opens up the possibility that TRIM72 interacts and ubiquitinates other viral proteins from various viruses, which warrants further investigation.This research contributes to our understanding of how TRIM72 and other TRIM proteins play a role in defending against viral invasions and may inspire further research in this field.

Ethics statement
Female C57 BL/6 mice (8-week-old) were purchased from the Hubei Center for Disease Control and Prevention, Hubei, China, and were housed in the animal facility at Huazhong Agricultural University in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Hubei Province, China.All experimental procedures involving animals were reviewed and approved by The Scientific Ethic Committee of Huazhong Agricultural University (permit number HZAUMO-2019-086).

Primary neuron cells
Primary neuron cells were isolated from the cerebral cortex of 16-day C57BL/6 mouse embryos as described previously [70].Briefly, the cerebral cortex was separated from the brains and then dissociated with 0.25% trypsin (wt/vol) at 37˚C for 30min.After centrifugation, the cells were resuspended in DMEM containing 5% FBS (Gibco) and subsequently passed through a 70 μm filter.After cell counting, the cells were seeded into 12-well plates (SORFA Life Science Co., Ltd, Beijing, China) that had been pretreated with poly-D-lysine and laminin (10 μg/ml) (Sigma, P7280).The cell supernatants were discarded at 5 h after seeding and replaced with Neurobasal plus Medium (Gibco, A3582901) containing 2% B-27 Plus (Gibco, A3582801).Then replace the medium every two days until 5-6 days post seeding for experiments.

qPCR analysis
Total RNA was isolated from cells and tissues using TRIzol reagent (Aidlab Biotech, Beijing, China, 321728AX) following the manufacturer's instruction.cDNAs were synthesized using HiScript II 1st Strand cDNA Synthesis Kit (Vazyme Biotech co., Ltd, R212-02), the genomic vRNA was synthesized using HiScript II Q RT SuperMix (Vazyme Biotech co.,ltd, R222-01) according to the manufacturer's protocol.qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd, Q711-02).The qPCR program consists of an initial denaturation step at 95˚C for 2 min for one cycle, followed by 40 cycles at 95˚C for 5 s and 60˚C for 30 s.The primer sequences used for qPCR were as follows (5'-

Indirect Immunofluorescence
N2a cells that transfected with pCAGGS or pCA-Trim72-flag for 48 h were fixed with 4% (v/v) paraformaldehyde for 15 min at room temperature then washed three times with PBS.The cells were permeabilized in PBS containing 0.3% Triton X-100 for 5 min at 4˚C, then blocked in 5% BSA buffer for 2 h at 37˚C.Cells were then incubated with anti-flag antibody (MBL, M185-3) for 1.5 h at 37˚C.After washed three times with PBS, cells were then incubated with Alexa Fluor 488 goat anti-mouse IgG (H+L) cross-adsorbed secondary antibody (Invitrogen, 11001) for 45 min at 37˚C.The cells were then stained with DAPI for 10 min and again washed three times with PBS, and then imaged with a NiKon microscope (Ti2-A).

Trim72 knockout N2a cell line construction
To construct a Trim72 -/-N2a cell line, the target sgRNA was designed as Trim72-sgRNA CCGTGCCTGCCTGATCCGGG, then the sgRNA sequence was cloned into px458-Cas9-puro plasmid (px458-Trim72-sgRNA-Cas9-puro).The recombinant plasmids were then transfected into N2a cells, and the positive cells were then separated with a flow cytometer and cultured which was used for the experiments.

PLOS PATHOGENS
TRIM72 lyssavirus assembly and/or release SM015010) or anti TRIM72 antibody (Proteintech, 22151-1-AP) which pre-conjugated with rProtein A/G magpoly beads was then added to the samples and incubated for 4 h at 4˚C.After incubation, the samples were washed 3 times with lysis buffer (5 min/time), and then the magnetic agarose was collected and resuspended in PBS.Then SDS-PAGE loading buffer was added to the samples and boiled for 10 min.The supernatants were collected and used for western blot analysis.

Fig 1 .
Fig 1. RABV infection induces up-regulation of TRIM72 in mouse brains and neuronal cells, and TRIM72 reduced RABV replication in neuronal cells.(A-B) 8-week-old female C57 BL/6 mice were intranasally infected with 100 fluorescence focus units (FFU) RABV (CVS-B2c) or mock-infected.The mouse brains were collected at 12 th d post-infection, the mRNA level of TRIM72 was analyzed by qPCR (A), and the protein levels of TRIM72 and RABV-N were analyzed by western blotting with anti-TRIM72 antibody and anti-RABV-N antibody (B).(C-D) Primary mouse neuron cells were isolated and cultured, then infected with RABV at different MOI for 36 h.The mRNA level of TRIM72 was analyzed by qPCR (C), and the protein levels of TRIM72 and RABV-N were analyzed by western blotting with anti-TRIM72 antibody and anti-RABV-N antibody (D).(E-F) N2a cells were infected with RABV at different MOI for 36 h.The mRNA level of TRIM72 was analyzed by qPCR (E), and the protein levels of TRIM72 and RABV-N were analyzed by western blotting with anti-TRIM72 antibody and anti-RABV-N antibody (F).(G) Empty vector (pCAGGS) or TRIM72-flag over-expression vector (pCA-Trim72-flag) were transfected into N2a cells respectively for 48 h, and then the TRIM72-flag level was analyzed by western blotting with anti-flag antibody.(H) Trim72 -/-N2a cells were constructed, then the TRIM72 level was analyzed in WT and Trim72 -/-N2a cells by western blotting with an anti-

Fig 3 .
Fig 3. TRIM72 restricts RABV infection at the viral release stage.(A) TRIM72-flag over-expression vector or empty vector were transfected into N2a cells respectively for 36 h, then infected with RABV (MOI = 0.1) for 1 h at 4˚C, the cells were collected and the viral genomic RNA (vRNA) levels were analyzed by qPCR.(B) TRIM72-flag over-expression vector or empty vector was transfected into N2a cells individually for 36 h then infected with RABV (MOI = 0.1) for 2 h at 37˚C, the cells were collected and the vRNA levels were analyzed by qPCR.(C-D) TRIM72-flag over-expression vector or empty vector were transfected into N2a cells individually for 12 h and then infected with RABV (MOI = 0.01) for the indicated time points, the cells were collected and the levels of RABV-N-mRNA (D) and vRNA (E) were analyzed by qPCR.(E-F) N2a cells were infected with rRABV, rRABV-Trim72, or rRABV-Trim72(-) at MOI 0.001.Then the cells were covered with culture medium containing 1% agar and incubated for 48 h.The viral spread was compared and the infected cell numbers were calculated within the fluorescence focus.Scale bar, 100 μm.(G) pCAGGS or pCA-Trim72-myc together with pCA-M-HA and pCA-G-flag were co-transfected into HEK-293T cells respectively for 48 h, and the protein levels of M-HA and G-flag in the supernatants and cells were analyzed by western blotting.Western blot data are representative of at least two independent experiments.https://doi.org/10.1371/journal.ppat.1011718.g003

Fig 4 .
Fig 4. TRIM72 induces proteasomal degradation of RABV-M by promoting K48-linked ubiquitination.(A) pCAGGS or pCA-Trim72-myc together with pCA-N, pCA-P, pCA-M-HA, and pCA-G-flag were co-transfected into N2a cells respectively for 48 h, and the protein levels of N, P, M-HA and G-flag were analyzed by western blotting.(B) pCAGGS or pCA-Trim72-flag were transfected into N2a cells for 12 h, then infected with RABV (MOI = 1) for 36h and RABV-N and RABV-M protein levels were analyzed by western blotting.(C) pCAGGS or pCA-Trim72-flag together with pCA-M-HA were co-transfected into N2a cells respectively.The specific inhibitors for proteasome and lysosome, Mg132 (10 μM) and NH 4 Cl (5 mM) were applied.Then the protein levels of TRIM72 and M were analyzed by western blotting.(D) pCA-M-HA together with pCAGGS or different concentrations of pCA-Trim72-flag were co- h and Co-IP assays were performed.As shown in Fig 5E, the TRIM72 truncations which contain the SPRY domain could interact with M, indicating that the TRIM72-SPRY domain directly interacts with M. Additionally, different M truncation over-expression vectors were also constructed based on its secondary structure (Fig 5F).Co-IP assays were performed between M truncations and TRIM72-SPRY domain.As shown in Fig 5G, All the M truncations including 49-160 aa were capable of interacting with the TRIM72-SPRY domain, while the M truncation with 49-160 aa deletion (i.e.1-48 and 161-202) cannot interact with the TRIM72-SPRY domain, indicating that M truncation (49-160 aa) was the key interaction domain that interacts with the TRIM72-SPRY domain.

Fig 5 .
Fig 5. TRIM72 directly interacts with RABV-M via the SPRY domain.(A) pCAGGS or pCA-Trim72-flag together with pCA-M-HA were co-transfected into N2a cells respectively.Then Mg132 (10 μM) was treated, and Co-IP assays were performed with anti-flag antibody posttransfection for 48 h.The protein levels of TRIM72-flag and M-HA were analyzed by western blotting.(B) pCAGGS or pCA-M-flag together with pCA-Trim72-HA were co-transfected into N2a cells respectively for 48 h.Then Co-IP assays were performed with anti-flag antibody.The protein levels of TRIM72-flag and M-HA were analyzed by western blotting.(C) N2a cells were infected with RABV (MOI = 1) for 36 h.

Fig 6 .
Fig 6.TRIM72 induces ubiquitination of RABV-M at K195. (A) Illustration of the lysine (K) sites in RABV-M.(B) The lysine residues in RABV-M were mutated to alanine and over-expression vectors were constructed and together with pCAGGS or pCA-Trim72-flag were co-transfected into N2a cells respectively for 48 h.The protein levels of M-HA mutations and TRIM72-flag were analyzed by western blotting.(C) WT M-HA or M-K195A-HA was over-expressed in N2a cells respectively for 48 h, and the protein level was analyzed by western blotting.(D) pCA-M-flag or pCA-M-K195A-flag together with UB-HA and pCAGGS or pCA-Trim72-myc were co-transfected into N2a cells and treated with Mg132 (10 μM), then Co-IP assays were performed with anti-flag antibody.The protein levels were analyzed by western blotting.(E) pCAGGS or pCA-Trim72-myc together with pCA-M-K195A-HA and pCA-G-flag were co-transfected into HEK-293T cells respectively for 48 h, the protein levels of M-K195A-HA and G-flag in the supernatants and cells were analyzed by western blotting.(F) Schematic diagram of the recombinant rRABV-M-K195A.(G) N2a cells were infected with rRABV or rRABV-M-K195A (MOI = 0.01) for the indicated time and their growth kinetics were compared.(H) pCAGGS or pCA-Trim72-flag was transfected into N2a cells respectively for 12 h, then infected with WT rRABV (MOI = 0.01) or rRABV-M-K195A (MOI = 0.01) for 48 h.The supernatants were collected and viral titers were