TRIM22-Mediated Apoptosis is Associated with Bak Oligomerization in Monocytes

Monocyte apoptosis is a key mechanism that orchestrates host immune responses during sepsis. TRIM22 is constitutively expressed at high levels in monocytes and plays important roles in the antiviral response and inflammation. Overexpression of TRIM22 interferes with the clonogenic growth of monocytic cells, suggesting that TRIM22 may regulate monocyte survival. However, the effect of TRIM22 on monocyte apoptosis remains unknown. In the present report, lipopolysaccharides (LPS)-primed human peripheral blood monocytes expressing higher levels of TRIM22 were more sensitive to apoptosis. This phenomenon was also observed in TRIM22-overexpressing THP-1 monocytes and was associated with the activation of caspase-9 and caspase-3, as well as the increased expression and oligomerization of the pro-apoptotic protein Bak. Similar expression patterns of TRIM22 and Bak were also observed in LPS-primed, apoptotic human peripheral blood monocytes. In addition, the deletion of either the RING domain or the SPRY domain of TRIM22 significantly attenuated TRIM22-mediated monocyte apoptosis and decreased Bak expression and oligomerization. Furthermore, in monocytes from septic patients, TRIM22 levels were down-regulated and positively correlated with Bak levels. Taken together, these results indicate that TRIM22 plays a critical role in monocyte apoptosis by regulating Bak oligomerization and may have a potential function in the pathogenesis of sepsis.

TRIM22 was first identified as an interferon-inducible protein that restricts HIV transcription. TRIM22 is constitutively expressed in peripheral blood leukocytes and lymphoid tissues, such as spleen and thymus 18 . The expression level of TRIM22 in monocytes was 2-fold higher than that in CD4 + and CD8 + T-lymphocytes and nearly 1.5-fold higher than that in B-lymphocytes 19 . TRIM22 is also a target gene of p53, which is a well-known regulator of cell growth and death. It has been noted that the overexpression of TRIM22 interferes with the clonogenic growth of monocytic U937 cells 20 , suggesting that it may participate in controlling monocyte survival. TRIM22 is also involved in host inflammatory responses 21 . Knockdown of TRIM30, the murine ortholog of TRIM22, increased the expression levels of pro-inflammatory cytokines (TNF-α , IL-6, IL-1β and IL-18). By contrast, the overexpression of TRIM30 protected mice against lipopolysaccharides (LPS)-induced septic shock 22,23 . However, whether TRIM30/TRIM22 affects monocyte apoptosis during sepsis remains unknown.
In the present study, the effect of TRIM22 on apoptosis was first investigated in human peripheral blood monocytes and the THP-1 monocytic cell line. We then identified the mechanism by which TRIM22 mediates apoptosis. We also observed how the structure of TRIM22 affects its function. Finally, we measured the expression levels of TRIM22 in septic patients and analyzed the correlations between TRIM22 and apoptosis-associated proteins.

Results
Increased endogenous TRIM22 levels in human peripheral blood monocytes were associated with cell apoptosis. We first investigated whether modulating the expression levels of endogenous TRIM22 could affect monocyte survival. TRIM22 transcription can be induced after LPS stimulation 24 . Here, we found that the protein levels of TRIM22 in human peripheral blood monocytes were upregulated > 1.5-fold upon LPS treatment (Fig. 1A). Under these conditions, the upregulation of TRIM22 did not affect monocyte apoptosis, as there were no differences in the proportions of Annexin V + cells between LPS-treated and untreated monocytes. However, when these cells were challenged with staurosporine (STS; an apoptosis inducer), we observed (A) Immunoblot analysis and densitometric quantification of TRIM22 expression in the lysates of human peripheral blood monocytes from three healthy donors in the presence or absence of LPS stimulation (100 ng/ml LPS for 16 h). **P < 0.01. (B) LPS-primed human peripheral blood monocytes were challenged with 0.5 μ g/ml STS for 8 h before harvest. The cells were stained with Annexin V-fluorescein isothiocyanate and propidium iodide for flow cytometry analysis. Histograms of annexin V + cells are shown, and quantitative data are presented as the mean ± SEM from 3 healthy volunteers. *P < 0.05. significantly more Annexin V + monocytes in LPS-primed cultures (47.2%) compared with unprimed cultures (40.5%; Fig. 1B). These findings show that monocytes with increased expression levels of TRIM22 are more susceptible to pro-apoptotic stimuli, suggesting a potential role for TRIM22 in mediating monocyte apoptosis.
Overexpression of TRIM22 sensitized THP-1 cells to apoptosis. To confirm our observations, we further investigated the role of TRIM22 in human peripheral blood monocyte apoptosis in vitro. To recapitulate the high levels of TRIM22 expression observed in human peripheral blood monocytes after LPS stimulation, a recombinant adenovirus, Ad.TRIM22, was used to overexpress TRIM22 in THP-1 cells. The expression levels of TRIM22 in infected THP-1 cells were about 1.5-fold of those in unstimulated peripheral blood monocytes while comparable to those in LPS-treated peripheral blood monocytes, and were similar to those in LPS-treated THP-1 cells ( Fig. 2A). As expected, when the cells were subjected to pro-apoptotic stimuli, TRIM22-overexpressing cells displayed increased levels of apoptosis compared with the control (Ad.LacZ-transduced) and mock-transduced cells ( Fig. 2B and C). TRIM22 mediated apoptosis through a caspase-9-dependent pathway. Next, we determined whether TRIM22-induced apoptosis was dependent upon caspase activity. In TRIM22-overexpressing cells, procaspase-3 and procaspase-9 were cleaved, and activated forms of caspase-3 and caspase-9 were elevated after STS challenge (Fig. 3A). STS-induced cytochrome c release was also increased in TRIM22-overexpressing cells (Fig. 3A). Moreover, when TRIM22-overexpressing cells were treated with the pan-caspase inhibitor Z-VAD-FMK, STS-induced apoptosis was completely abolished ( Fig. 3B and C). These findings demonstrate that TRIM22 sensitizes monocyte to apoptosis in a caspase-dependent manner.
Overexpression of TRIM22 modulated Bak expression and oligomerization. We next investigated whether TRIM22-mediated apoptosis induced changes in other proteins involved in the intrinsic apoptosis pathway. The basal expression levels of Bcl-2 in TRIM22-overexpressing THP-1 cells were lower than those in mockand Ad.LacZ-transduced cells, but were not further suppressed upon STS stimulation. Interestingly, we found that the expression levels of Bak were significantly upregulated following STS treatment in TRIM22-overexpressing cells (Fig. 4A). This upregulation of Bak was not associated with changes in the half-life of Bak because the inhibition of transcription or translation did not affect levels of Bak expression (Fig. 4B). Bak forms large oligomeric complexes that trigger cytochrome c release from the mitochondria 25,26 . After the induction of apoptosis, we observed more Bak oligomers in TRIM22-overexpressing THP-1 cells (Fig. 4C). To examine whether TRIM22 promoted Bak oligomerization directly or was dependent on increased Bak protein synthesis during apoptosis, we evaluated the oligomerization in apoptotic cells pretreated with the protein synthesis inhibitor cycloheximide. After treatment, increased Bak oligomerization in TRIM22-overexpressing cells was still observed (Fig. 4D), demonstrating that TRIM22 induces Bak oligomerization.
The relationship between TRIM22 and Bak was further studied in LPS-primed, STS-challenged peripheral blood monocytes from healthy volunteers. Correlation analysis showed a positive correlation between the expression levels of TRIM22 and Bak (r = 0.534, P < 0.0001) (Fig. 4E). Moreover, upon STS challenge, higher Bak protein levels were observed in LPS-primed peripheral blood monocytes, which have already been shown to demonstrate inducible TRIM22 expression (Fig. 4F). These findings further demonstrate a critical role for the pro-apoptotic protein Bak in TRIM22-mediated apoptosis.
RING and SPRY domains were involved in TRIM22-mediated monocyte apoptosis. Previous studies have shown that the RING and SPRY domains play key roles in TRIM22 function 27,28 . To determine whether these domains were involved in the pro-apoptotic role of TRIM22, we constructed recombinant adenoviruses expressing domain-deletion mutants, including TRIM22-Δ RING and TRIM22-Δ SPRY ( Fig. 5A and B), and we evaluated their effects on STS-induced apoptosis. Deletion of either the RING or SPRY domain moderately increased cell apoptosis ( Fig. 5C and D), and partially blocked Bak expression (Fig. 5E) and oligomerization (Fig. 5F). Furthermore, oligomerization of Bak was not influenced by cycloheximide treatment (Fig. 5G).
TRIM22 expression levels positively correlated with Bak expression in monocytes from septic patients. We further measured the expression levels of TRIM22 and Bak in peripheral blood monocytes collected from septic patients. The demographic and clinical characteristics of enrolled septic patients are listed in Table 1. The septic group consisted of 15 patients, including five patients with pneumonia, four patients with hepatobiliary infection, two patients with peritonitis, two patients with pancreatitis, and two patients with soft-tissue infection. Infection was documented in all patients by microbiologic inspection. The control subjects consisted of 8 non-septic, critically ill surgical patients (age 54.1 ± 17.7 years; 4 females, 4 males). Real-time quantitative PCR analysis showed that TRIM22 mRNA levels in the septic patients were significantly lower compared with those in the non-septic controls (Fig. 6A). Furthermore, TRIM22 mRNA levels were positively correlated with Bak levels in septic patients (r = 0.6064, P = 0.0006; Fig. 6B). Since Bak is an important pro-apoptotic protein, these findings indicate that in sepsis, monocyte survival may be autoregulated via the control of Bak-associated TRIM22 expression.

Discussion
In this study, increasing the levels of both endogenous and exogenous TRIM22 sensitized monocytes to STS-induced apoptosis. This function of TRIM22 was related to the increased expression and oligomerization of Bak via a caspase-dependent pathway and was associated with the RING and SPRY domains of the TRIM22 molecule. In addition, the mRNA levels of TRIM22 were down-regulated and positively correlated with Bak transcripts in monocytes from septic patients.
As a p53 target gene, TRIM22 inhibits the clonogenic growth of U937 monocytic cells 20 . In this study, we found that LPS-primed monocytes expressing high levels of TRIM22 were more sensitive to STS-induced apoptosis. In addition, the recombinant adenovirus-mediated overexpression of TRIM22 enhanced the susceptibility of monocytes to STS-induced apoptosis. In light of previous studies regarding the role of TRIM22 in antiviral immunity, cytokine production, and inflammatory diseases, our findings suggest that TRIM22 may mediate inflammation by controlling monocyte survival.
Previous studies have shown that TRIM19 is required for the activation of caspase-1 and caspase-3 in mouse splenocytes, suggesting that TRIM19 is involved in caspase-dependent apoptosis 13 . By contrast, the overexpression of TRIM19 induced apoptosis in rat embryonic fibroblasts in the absence of caspase-3 activation, and the caspase inhibitor Z-VAD-FMK failed to block TRIM19-induced cell death 14 . These data reveal that TRIM proteins can participate in apoptosis via caspase-dependent or -independent pathways, potentially in a cell type-specific manner. In our study, TRIM22 overexpression promoted the cleavage of procaspase-9 and procaspase-3, elevated the expression levels of cleaved caspase-3 and cleaved caspase-9, and enhanced cytochrome c release in STS-challenged monocytic cells. Moreover, pretreatment with the caspase inhibitor Z-VAD-FMK effectively inhibited TRIM22-mediated apoptosis. Together, these findings demonstrate that TRIM22 promotes monocyte apoptosis via a caspase-dependent pathway.
p53 is an important regulator of cell growth suppression and apoptosis. p53 induces apoptosis by regulating the transcription of pro-apoptotic and anti-apoptotic genes such as Bax and Bcl-2 29 . Given that TRIM22 is a target gene of p53, we propose that TRIM22 promotes monocyte apoptosis by regulating the Bcl-2 family proteins. Following STS challenge, the overexpression of TRIM22 significantly enhanced Bak expression but did not affect the expression of other Bcl-2 family proteins such as Bax and Bcl-xl. This initial finding was further confirmed in STS-challenged LPS-primed human peripheral blood monocytes, in which both mRNA and protein levels of Bak were positively correlated with those of TRIM22. These data suggest a critical role for Bak in TRIM22-sensitized apoptosis.
In stressed cells, inactive Bak undergoes an activating conformational change leading to the formation of higher-order multimers, followed by oligomerization. Bak oligomerization enhances the permeabilization of the outer mitochondrial membrane, which results in the release of pro-apoptogenic factors (such as cytochrome c) from the mitochondria into the cytosol 26  . The multimer conformation of Bak was visualized by immunoblotting. VDAC1 was used as a mitochondrial loading control. (D) Cells were pre-treated with CHX (20 μ g/ml) for 1 h to block Bak protein synthesis. Oligomerization of Bak was examined as indicated. Non cross-linked Bak incubated with DMSO control buffer was run as a monomer. (E) Monocytes isolated from 23 healthy volunteers were exposed to 100 ng/ml LPS for 16 h, treated with 0.5 μ g/ml STS for 8 h before harvest. Correlations between the expression levels of TRIM22 and Bak mRNA were analyzed (r = 0.534, P < 0.0001). (F) Representative immunoblots and densitometric quantifications illustrating protein levels of TRIM22 and Bak in monocytes. TRIM22 and Bak levels were normalized to β -actin. Data are presented as the mean ± SEM from three healthy volunteers. *P < 0.05.
Scientific RepoRts | 7:39961 | DOI: 10.1038/srep39961 protein in TRIM22-overexpressing monocytes independent of protein synthesis, suggesting a role for TRIM22 in apoptosis-associated Bak oligomerization. This may also explain the sensitization to apoptosis in LPS-treated monocytes expressing higher levels of TRIM22. However, whether TRIM22-mediated apoptosis was caused by Bak oligomerization or triggered via other pathways remains unknown and requires further investigation.
TRIM22 contains a conserved RBCC structure beginning with a RING domain at the N-terminus and followed by a B30.2/SPRY domain at the C-terminus 33 . Previous studies have demonstrated that these distinct domains mediate the diverse functions of TRIM22. The RING domain is important for the E3 ubiquitin ligase activity of TRIM proteins 34 , which is associated with the effects of TRIM family members on cell survival via the ubiquitination and proteasomal degradation of p53 or other apoptosis-related proteins 12,16,[35][36][37] . The function of the SPRY domain is not well understood, although several studies suggest it may mediate protein-protein interactions 38,39 . Both the RING and SPRY domains of TRIM22 are essential for TRIM22-mediated anti-HBV activity and the activation of NF-κ B 40,41 . In this study, using domain-deletion mutants, we found that deletion of either the RING domain or the SPRY domain significantly attenuated STS-induced apoptosis, which was associated with decreased Bak expression and oligomerization. However, what would both domains being so different only partially and equally contribute to change Bak expression and oligomerization or whether the observations are attributed the truncated peptides but not the specific domain truncated remians unclear. Further additional experiments to dissect these mechanisms should help us understand well. After 72 h, the expression levels of the relevant proteins were measured using Western blots. (C) Mock-infected cells or cells infected with wild-type TRIM22, domain-deletion mutants or control adenoviruses were treated with 0.5 μ g/ml STS for 4 h, and apoptosis was analyzed using flow cytometry. (D) Histograms of annexin V + cells are shown, and quantitative data are presented as the mean ± SEM from three independent experiments. *P < 0.05. The expression levels (E) and oligomerization of Bak (F) were measured using Western blots. VDAC1 was used as a mitochondrial loading control. (G) THP-1 cells were pre-treated with 20 μ g/ml CHX for 1 h and oligomerization of Bak was analyzed as described above.
Scientific RepoRts | 7:39961 | DOI: 10.1038/srep39961 Sepsis induces a multitude of defects in immunity that causes aberrant inflammation, immune suppression, susceptibility to infections, and death. One of the manifestations of sepsis-induced immunosuppression is monocyte/macrophage dysfunction 42 . Monocytes/macrophages are important players in the pathogenesis of sepsis. Monocytes/macrophages from septic patients undergo functional reprogramming from a proinflammatory to an immunosuppressive phenotype 43 . The proinflammatory response often predominates in the early phase of an infection. And most patients will rapidly progress to an immunosuppressive state, characterized by decreased phagocytic ability, reduced bactericidal activity, and attenuated proinflammatory cytokine production 42 . As TRIM22 expression was induced upon LPS challenge in human peripheral blood monocytes from healthy donors, decreased TRIM22 levels in septic patients might result from immunosuppression.
Previous studies have demonstrated that the overexpression of TRIM30, the mouse ortholog of TRIM22, protected mice from LPS-induced septic shock 22,23 . Although these studies did not focus on the effects of TRIM30 on monocyte apoptosis, given the present findings, it is reasonable to speculate that TRIM30 might also improve outcomes in septic mice through the sensitization of monocytes to apoptosis.
In conclusion, we show that TRIM22 sensitizes monocytes to STS-induced apoptosis. Upregulation of TRIM22 triggers the expression and oligomerization of Bak and subsequently leads to cytochrome c release in a caspase-9-and caspase-3-dependent manner. Both the RING domain and the SPRY domain of the TRIM22 molecule are associated with its pro-apoptotic function. These findings not only illustrate the role of TRIM22 in monocyte apoptosis but also indicate the potential functions of TRIM22 in inflammatory diseases such as sepsis.   polyvinylidene difluoride membranes (Millipore, Billerica, Massachusetts, USA). The membranes were blocked with 5% skim milk in Tris-buffered saline with 0.05% Tween-20 for 1 h at room temperature, incubated overnight with specific primary antibodies, washed three times with Tris-buffered saline containing 0.05% Tween-20, and further incubated for 1 h with appropriate horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, Pennsylvania, USA). After washing the membranes with Tris-buffered saline containing 0.05% Tween-20, the protein bands were visualized with an EZ-ECL kit (Bioind, Kibbutz, Beit Haemek, Israel). The rabbit-derived primary antibodies included Bcl-2, Bcl-xL, Bak, Bax, and cytochrome c (all from Epitomics, Inc., Burlingame, California, USA), as well as cleaved caspase-3 and caspase 9 (Cell Signaling Technology, Inc., Beverly, Massachusetts, USA). A mouse anti-β -actin monoclonal antibody (Sigma, St. Louis, Missouri, USA) was used as a loading control.
Cross-linking. The Bak oligomerization assay was performed as previously reported 47 . Briefly, mitochondria were isolated and incubated with 4 mM disuccinimidyl suberate (Sigma, St. Louis, Missouri, USA) for 30 min at room temperature. Cross-linked samples were analyzed by western blotting using an anti-Bak antibody (Epitomics, Burlingame, California, USA). Rabbit anti-VDAC1 monoclonal antibodies (Abcam, Cambridge, Massachusetts, USA) were used as a mitochondrial loading control.
Statistical analysis. Data are presented as the mean ± SEM. Statistical significance among groups was assessed by One-way ANOVA using GraphPad Prism 5.0 (GraphPad Software Inc., La Jolla, California, USA). Bonferroni's test was used to correct for multiple comparisons where applicable. The relationship between the expression levels of TRIM22 and Bak was assessed using the Spearman correlation test. Differences were considered statistically significant when a two-tailed P value was less than 0.05.