Visual Abstract

Salmonella enterica serovar Typhimurium (S. Typhimurium) is a Gram-negative bacterium that induces cell death of macrophages as a key virulence strategy. We have previously demonstrated that the induction of macrophage death is dependent on the host’s type I IFN (IFN-I) response. IFN-I signaling has been shown to induce tripartite motif (TRIM) 21, an E3 ubiquitin ligase with critical functions in autoimmune disease and antiviral immunity. However, the importance and regulation of TRIM21 during bacterial infection remains poorly understood. In this study, we investigated the role of TRIM21 upon S. Typhimurium infection of murine bone marrow–derived macrophages. Although Trim21 expression was induced in an IFN-I–dependent manner, we found that TRIM21 levels were mainly regulated posttranscriptionally. Following TLR4 activation, TRIM21 was transiently degraded via the lysosomal pathway by chaperone-mediated autophagy (CMA). However, S. Typhimurium–induced mTORC2 signaling led to phosphorylation of Akt at S473, which subsequently impaired TRIM21 degradation by attenuating CMA. Elevated TRIM21 levels promoted macrophage death associated with reduced transcription of NF erythroid 2–related factor 2 (NRF2)–dependent antioxidative genes. Collectively, our results identify IFN-I–inducible TRIM21 as a negative regulator of innate immune responses to S. Typhimurium and a previously unrecognized substrate of CMA. To our knowledge, this is the first study reporting that a member of the TRIM family is degraded by the lysosomal pathway.

The Gram-negative bacterium Salmonella enterica serovar Typhimurium (S. Typhimurium) is a facultative intracellular pathogen that causes severe gastroenteritis in humans and typhoid fever in mice (1). Once S. Typhimurium has passed the intestinal epithelial barrier, it is sensed and targeted by cells of the innate immune system, predominantly macrophages (1). However, S. Typhimurium has evolved several strategies to survive and replicate within macrophages. Intracellular survival of S. Typhimurium critically depends on the expression of bacterial effector proteins, which are encoded by the Salmonella pathogenicity island (SPI)–2 (2). We have previously demonstrated that S. Typhimurium uses its SPI-2 genes to interfere with cellular energy homeostasis. During S. Typhimurium infection of macrophages, the SPI-2–encoded effector protein SsrB promoted lysosomal degradation of the host’s energy sensor AMP kinase (AMPK), resulting in impaired autophagy (3).

Macroautophagy (hereafter, autophagy) is a highly conserved process of cellular self-digestion by which protein aggregates or damaged organelles are engulfed into LC3 positive double-membrane vesicles called autophagosomes. Autophagosomes subsequently fuse with lysosomes, leading to the enzymatic degradation of endogenous cargo (4, 5). Although activation of AMPK and inhibition of mammalian target of rapamycin complex (mTORC) 1 induce autophagy, activation of Akt and mTORC2 impair lysosomal activity and thus autophagic degradation (6, 7). Proteins designated for autophagic degradation are marked with a ubiquitin signal that recruits an autophagic adaptor protein, such as sequestome 1 (hereafter, p62) (8). p62 simultaneously interacts with ubiquitin via its ubiquitin-associated (UBA) domain and delivers (poly)ubiquitinated cargo either to the autophagosome via its LC3-interacting region (LIR) or to the proteasome via its Phox-BEM1 (PB1) domain (9, 10).

Tripartite motif (TRIM) 21, also known as Ro52/SS-A, is an E3 ubiquitin ligase first discovered as an autoantigen in patients with systemic lupus erythematosus and Sjögren syndrome (11, 12). Beyond its role in autoimmune diseases, TRIM21 has important functions in the innate immune defense to various pathogens as it catalyzes the formation of ubiquitin chains on numerous substrate proteins involved in the antimicrobial response (13, 14). Proteins decorated with Lys48 (K48)–linked ubiquitin chains usually undergo proteasomal degradation (15, 16), whereas Lys63 (K63)–linked ubiquitin chains modify protein trafficking and function as well as regulate inflammatory signaling cascades, which often culminate in the activation of the transcription factor NF-κB (17, 18). Recently, TRIM21 has been implicated to regulate cellular redox homeostasis through ubiquitination and proteasomal degradation of p62 (16). Reduced p62 levels promoted cytosolic interaction of Kelch-like ECH-associated protein 1 (KEAP1) with NF erythroid 2–related factor 2 (NRF2), a major regulator of antioxidative stress responses, resulting in impaired transcription of antioxidative genes (16). Like several other members of the TRIM family, TRIM21 has been suggested to balance inflammatory responses through ubiquitination of IFN regulatory factors (IRFs) and, thus, type I IFN (IFN-I) induction (14, 19, 20). Because the activation of IFN-I signaling is one of the hallmarks of antiviral immunity, most of our knowledge on TRIM21/IFN-I regulation is derived from studies investigating viral infections (15, 21).

In recent years, however, many studies have highlighted the significance of IFN-I signaling also during bacterial infection, in which it can have both beneficial and detrimental effects (2224). We have previously reported that S. Typhimurium exploits the host’s IFN-I response to induce cell death of macrophages (25). Mechanistically, IFN-I signaling impaired the host’s ability to respond to S. Typhimurium–induced oxidative damage through scavenging NRF2 in the cytosol (26). As a consequence of reduced macrophage numbers, infected mice had reduced ability to control S. Typhimurium infection, resulting in enhanced bacterial dissemination and increased mortality (25). Therefore, macrophage survival upon S. Typhimurium infection is thought to be protective for the host.

Because TRIM21 has been associated with the regulation of IFN-I responses (14) and our previous work has underscored the importance of IFN-I signaling during S. Typhimurium infection (25, 26), we sought to investigate the function of TRIM21 during S. Typhimurium infection in more detail. We report in this study that S. Typhimurium infection induces Trim21 expression in an IFN-I–dependent manner and that high levels of TRIM21 sensitize S. Typhimurium–infected macrophages to cell death. Specifically, we show that TRIM21 is transiently degraded through the lysosomal pathway by chaperone-mediated autophagy (CMA). However, S. Typhimurium induces mTORC2/Akt signaling to impair CMA, resulting in increased TRIM21 levels and subsequent macrophage death.

Bone marrows were isolated from the femurs of wild-type (WT) C57BL/6, IFN-I receptor (Ifnar1−/−), autophagy-related protein 7 (Atg7−/−), myeloid differentiation primary response 88 (MyD88−/−), and TIR domain–containing adapter-inducing IFN-β (TRIF−/−)–deficient mice after cervical dislocation. Bone marrow was collected in RPMI medium (Merck Millipore, Burlington, MA) containing 10% FBS (Life Technologies, Thermo Fisher Scientific, Waltham, MA) and 10% DMSO (Sigma-Aldrich, St. Louis, MO) as described previously (25) and stored in liquid nitrogen until further usage. All animal procedures were performed according to the institutional guidelines on animal welfare and were approved by the North Rhine-Westphalia State Agency for Nature, Environment and Consumer Protection (Landesamt für Natur, Umwelt and Verbraucherschutz Nordrhein-Westfalen; file no. 84-02.05.40.14.082 and 84-02.04.2015.A443) and the local animal care committee (University of Cologne).

Bone marrows were differentiated into macrophages in RPMI medium supplemented with 10% FBS and 20% L929 cell–culture supernatant for 7 d. Twenty-four hours prior to infection, bone marrow–derived macrophages (BMDMs) were seeded into 6-well or 24-well tissue culture plates (Techno Plastic Products, Trasadingen, Switzerland), respectively, and adherent BMDMs were used for experiments from day 7 to 9.

Overnight cultures of S. Typhimurium (SL1344), S. Typhimurium ΔSPI-1 and ΔsopA mutant strains were transferred into fresh brain–heart infusion (BHI) broth (Oxoid, Thermo Fisher Scientific) and were grown to late-exponential phase at 37°C with constant agitation. BMDMs were infected with the specified bacteria at a multiplicity of infection of 10. Infected BMDMs were incubated at 37°C for 30 min to ensure bacterial uptake. Next, BMDMs were washed with RPMI medium containing 50 μg/ml gentamicin (Thermo Fisher Scientific) to remove extracellular bacteria. Infected BMDMs were then incubated for up to 2 h in RPMI medium containing 10% FBS and 50 μg/ml gentamicin before the medium was exchanged to RPMI medium containing 10% FBS and 10 μg/ml gentamicin. Where indicated, noninfected BMDMs were treated with S. Typhimurium–derived LPS (200 ng/ml; Sigma-Aldrich) diluted in RPMI medium supplemented with 10% FBS and 50 or 10 μg/ml gentamicin, respectively. At the indicated time points, BMDMs were washed once with PBS and collected for experiments.

Two hours prior to infection, BMDMs were incubated with one of the following inhibitors diluted in RPMI medium containing 10% FBS: MG132 (50 μM; Selleck Chemicals, Houston, TX), Concanamycin A (100 nM; Sigma-Aldrich), Akt inhibitor VIII (10 μM; Enzo Life Sciences, Farmingdale, NY), Torin1 (10 μM; Tocris Bioscience, R&D Systems, Minneapolis, MN), and rapamycin (0.5 μM; Selleck Chemicals). Next, BMDMs were infected with S. Typhimurium (SL1344) as described. At 2 h of infection, each inhibitor was diluted to half of the initial concentration in RPMI medium supplemented with 10% FBS and 10 μg/ml gentamicin. At the indicated time points, BMDMs were washed once with PBS and collected for experiments.

S. Typhimurium (SL1344) was grown to late-exponential phase in BHI medium as outlined above. Bacteria were pelleted at 14,000 rpm for 5 min at 4°C and resuspended either in 1 ml of RPMI medium containing 10% mouse serum (Sigma-Aldrich) or RPMI medium only. Bacteria were left at room temperature for 30 min to allow binding of serum Abs. Next, BMDMs were infected with either opsonized or nonopsonized S. Typhimurium at a multiplicity of infection of 10.

BMDMs (1.0 × 106 cells per well) were lysed with radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitor mixture (Thermo Fisher Scientific). Total protein concentrations were estimated using the BCA Protein Assay Kit (Pierce, Thermo Fisher Scientific), and equal amounts of protein were loaded and separated on a 10% SDS-PAGE gel. Proteins were transferred to a polyvinylidene difluoride membrane (GE Healthcare, Chalfont St. Giles, U.K.), followed by blocking with 5% milk or 5% BSA, respectively, diluted in TBST (0.05% Tween 20) for 1 h at room temperature. Next, membranes were incubated overnight at 4°C with one of the following primary Abs: TRIM21 (no. ab207728; Abcam), SQSTM1/p62 (no. 5114; Cell Signaling Technology), NRF2 (no. sc-722; Santa Cruz Biotechnology), β-actin (no. sc-47778; Santa Cruz Biotechnology), LC3 (no. L7543; Sigma-Aldrich), Rictor (no. 2140; Cell Signaling Technology), and phospho-Akt serine 473 (no. 4060; Cell Signaling Technology). Membranes were washed with TBST and incubated with an appropriate secondary HRP-conjugated Ab (no. HAF007 and no. HAF008; R&D Systems) for 1 h at room temperature. After washing, membranes were treated with an ECL substrate (GE Healthcare) for 1 min and exposed to x-ray film (Advansta, San Jose, CA). Densitometric quantification of signal intensity was performed using ImageJ 1× software (National Institutes of Health) (27).

BMDMs (0.1 × 106 cells per well) seeded on glass coverslips were infected with S. Typhimurium and fixed with 4% (w/v) formaldehyde in PBS for 15 min at room temperature. Where indicated, BMDMs were stained with 250 nM LysoTracker Deep Red (no. L12492; Invitrogen, Thermo Fisher Scientific) according to the manufacturer’s protocol prior to fixation. Next, BMDMs were permeabilized for 5 min with 0.3% Triton X-100 in PBS, and samples were incubated for 30 min with Image-iT R FX signal enhancer (Invitrogen, Thermo Fisher Scientific). After blocking with 10% FBS and 0.03% Triton X-100 diluted in PBS for 1 h, samples were incubated overnight at 4°C with one of the following primary Abs: TRIM21 (no. sc-21367; Santa Cruz Biotechnology), S. Typhimurium LPS (no. MA1-83451; Thermo Fisher Scientific), lysosome-associated membrane protein type 2A (LAMP2A) (no. ab18528; Abcam), or HSC70 (no. ab19136; Abcam). Samples were then washed with 0.03% Triton X-100 diluted in PBS and incubated with an appropriate fluorescent secondary Ab (no. A11006, no. A11008, no. A11072, and no. A31553; Thermo Fisher Scientific) for 1 h at room temperature protected from light. Where indicated, coverslips were mounted on glass slides using ProLong Gold antifade containing DAPI (no. P36935; Invitrogen, Thermo Fisher Scientific). Images were acquired with a confocal Leica TCS SP8-X microscope equipped with a white light laser, a 405-nm diode UV laser, and a ×100 objective lens (HCX Plan-Apochromat CS ×100 oil, 1.46 numerical aperture). Fiji software (National Institutes of Health) was used for image acquisition (28).

Colocalization was analyzed from confocal images using Fiji’s ImageJ Colocalization Test plugin (28) with Fay randomization to calculate Pearson correlation coefficient for channel 1 (red) and channel 2 (green). This value was compared with what would be expected for random overlap. The observed Pearson correlation coefficient was considered significant if it was >95% of the correlations between channel 1 and a number of randomized channel 2 images. Pearson correlation was determined across 50 cells from three independent experiments.

BMDMs (1.0 × 106 cells per well) were transfected for 48 h with either 50 nM of nontargeting small interfering RNA (siRNA) (no. SR-CL000-005; Eurogentec, Fawley, U.K.) or siRNA specific for Trim21 (no. L-060414-00-0005; Dharmacon, Horizon, Lafayette, CO) or Rictor (no. L-064598-01-0005; Dharmacon, Horizon) using GenMute siRNA Transfection Reagent (no. SL100568; SignaGen Laboratories, Rockville, MD) according to the manufacturer’s instructions.

Cell viability was assessed by trypan blue staining as described previously (16). Briefly, BMDMs (1.0 × 106 cells per well) were pelleted at 14,000 rpm for 5 min at 4°C at the indicated time points, and cells were resuspended in 250 μl of RPMI medium supplemented with 10% FBS. Next, 50 μl of the cell suspension was diluted with an equal amount of 0.4% trypan blue (Invitrogen, Thermo Fisher Scientific) and incubated at room temperature for up to 3 min. Ten microliters of the trypan blue cell suspension was transferred to a counting chamber (Countess Cell Counting Chamber Slides; Invitrogen, Thermo Fisher Scientific) and the relative and total amounts of live cells were determined from quintuplicates using an automated cell counter (Countess II Automated Cell Counter; Invitrogen, Thermo Fisher Scientific) according to the manufacturer’s instructions.

After 0 and 6 h of infection with serum opsonized S. Typhimurium, BMDMs (0.5 × 106 cells per well) were lysed with 1% Triton X-100 and 0.01% SDS/PBS solution. Lysates from three biological replicates per time point were serially diluted with sterile PBS and 100 μl were spread on BHI agar plates. BHI plates were incubated overnight at 37°C and CFU were counted the next day to determine bacterial uptake (0 h) and intracellular survival (6 h).

The RNeasy Mini Kit (QIAGEN, Venlo, the Netherlands) was used for total RNA extraction from BMDMs (1.0 × 106 cells per well), and cDNA (500 ng) was synthesized with random hexamers using SuperScript III reverse transcriptase (Invitrogen, Thermo Fisher Scientific). Primer sequences were designed using Primer3 software and Basic Local Alignment Search Tool (National Center for Biotechnology Information, Bethesda, MD) and are listed in Table I. All primers were purchased from Invitrogen. A total of 25 μl was used for PCR, each sample containing 10 ng of cDNA, 0.4 μmol/l of each forward and reverse primer, and master mixture (SsoFast EvaGreen Supermix; Bio-Rad Laboratories, Hercules, CA). Real-time PCR was carried out on a Bio-Rad Laboratories platform (PCR Cycler CFX96; Bio-Rad Laboratories) as follows: initial denaturation step at 95°C for 2 min, 40 cycles at 95°C for 5 s, and 60°C for 15 s, followed by a denaturation step at 95°C for 60 s. Melt curve analysis was performed to check amplification specificity. Relative mRNA expression was analyzed using the comparative threshold cycle method with hypoxanthine-guanine phosphoribosyltransferase (Hprt) as the endogenous reference gene for all reactions. The relative mRNA expression of uninfected BMDMs was used for normalization of infected BMDMs. All assays were performed in biological triplicates, and a nontemplate control was included in every experiment to exclude DNA contamination.

Statistical analyses were performed using GraphPad Prism v5 software. Differences between groups were assessed by two-tailed unpaired Student t test or one-way ANOVA with repeated measures when more than two groups were analyzed. Data are presented as mean ± SD. Differences were considered statistically significant when p ≤ 0.05 (*), very significant when p ≤ 0.01 (**), and highly significant when p ≤ 0.001 (***). All experiments were repeated at least three times, and statistical analyses were performed on three independent experiments.

To investigate whether TRIM21 expression is regulated by IFN-I signaling, we infected WT and IFN-IR receptor–deficient (Ifnar1−/−) BMDMs with S. Typhimurium and analyzed relative mRNA expression of Trim21. As shown in Fig. 1A, S. Typhimurium infection led to a strong upregulation of Trim21 in WT BMDMs 1 and 6 h postinfection (p.i.), respectively, whereas Trim21 expression was strongly reduced in Ifnar1−/− BMDMs. Similarly, we detected significantly more TRIM21 in total cell lysates of S. Typhimurium–infected WT BMDMs compared with Ifnar1−/− BMDMs, in which TRIM21 was hardly detectable throughout infection (Fig. 1B, 1C). These results demonstrate that TRIM21 is expressed during S. Typhimurium infection of macrophages and that its expression is dependent on the IFN-I signaling pathway.

FIGURE 1.

TRIM21 is induced in an IFN-I–dependent manner. (A) WT and Ifnar1−/− BMDMs were infected with S. Typhimurium for 1 and 6 h, respectively, and relative mRNA expression levels of Trim21 were determined by quantitative real-time PCR (qRT-PCR). Values were normalized to the amounts of mRNA of uninfected WT BMDMs. (B) Total cell lysates of WT and Ifnar1−/− BMDMs infected with S. Typhimurium for 1, 4, or 6 h were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin was used as a loading control. (C) TRIM21 bands from three independent immunoblots were quantified densitometrically and signal intensities were normalized to the corresponding β-actin intensities. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

FIGURE 1.

TRIM21 is induced in an IFN-I–dependent manner. (A) WT and Ifnar1−/− BMDMs were infected with S. Typhimurium for 1 and 6 h, respectively, and relative mRNA expression levels of Trim21 were determined by quantitative real-time PCR (qRT-PCR). Values were normalized to the amounts of mRNA of uninfected WT BMDMs. (B) Total cell lysates of WT and Ifnar1−/− BMDMs infected with S. Typhimurium for 1, 4, or 6 h were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin was used as a loading control. (C) TRIM21 bands from three independent immunoblots were quantified densitometrically and signal intensities were normalized to the corresponding β-actin intensities. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

Close modal

Because we had previously shown that IFN-I signaling abrogates the NRF2-dependent antioxidative response during S. Typhimurium infection and TRIM21 has been reported as a negative regulator of NRF2, we investigated the role of IFN-I–inducible TRIM21 during S. Typhimurium–induced cell death in more detail.

Table I.
Primer sequences for qRT-PCR used in this study
GeneSequence
Hmox-1 (for) 5′-GGTCAGGTGTCCAGAGAAGG-3′ 
Hmox-1 (rev) 5′-CTTCCAGGGCCGTGTAGATA-3′ 
Gclc (for) 5′-GTGGACGAGTGCAGCAAG-3′ 
Gclc (rev) 5′-GTCCAGGAAATACCCCTTCC-3′ 
Hprt (for) 5′-GTTGGATACAGGCCAGACTTTGTTG-3′ 
Hprt (rev) 5′-GATTCAACTTGCGCTCATCTTAGGC-3′ 
GeneSequence
Hmox-1 (for) 5′-GGTCAGGTGTCCAGAGAAGG-3′ 
Hmox-1 (rev) 5′-CTTCCAGGGCCGTGTAGATA-3′ 
Gclc (for) 5′-GTGGACGAGTGCAGCAAG-3′ 
Gclc (rev) 5′-GTCCAGGAAATACCCCTTCC-3′ 
Hprt (for) 5′-GTTGGATACAGGCCAGACTTTGTTG-3′ 
Hprt (rev) 5′-GATTCAACTTGCGCTCATCTTAGGC-3′ 

for, forward; qRT-PCR, quantitative real-time PCR; rev, reverse.

In accordance with our previous work (25, 26), S. Typhimurium–induced cell death of macrophages after 1 and 6 h of infection (Fig. 2A). Cell death was partially mediated by TRIM21 because Trim21 knockdown significantly improved macrophage viability (Fig. 2A). We hypothesized that TRIM21 induced cell death by interfering with the p62/NRF2 pathway because knockdown of TRIM21 resulted in a modest increase of p62 at 6 h p.i. (Fig. 2B–D). It should be noted, however, that the increase in p62 as quantified in Fig. 2D may not precisely reflect p62 band intensities of the immunoblot shown in Fig. 2B because of differences in TRIM21 knockdown efficiency in various experiments (ranging from ∼50–80%). Importantly, although S. Typhimurium decreased NRF2 expression at 1 and 6 h p.i., the corresponding NRF2 levels were higher following silencing of Trim21 (Fig. 2B, 2E). Consequently, knockdown of TRIM21 significantly upregulated the NRF2-dependent antioxidative genes Gclc and Hmox-1 at 1 and 6 h p.i. (Fig. 2F, 2G). Unlike its role in mediating macrophage death, TRIM21 was dispensable for controlling intracellular survival of S. Typhimurium, as the numbers of live bacteria were similar in BMDMs transfected with nontargeting small interfering RNA (siCtrl) or small interfering RNA against Trim21 (siTrim21) (Supplemental Fig. 1). We therefore conclude that TRIM21 contributes to S. Typhimurium–induced cell death by interfering with p62/NRF2-dependent antioxidative stress responses, which occurs independent of bacterial survival.

FIGURE 2.

TRIM21 promotes cell death. (A) WT BMDMs were transfected with nontargeting siRNA (siCtrl) or siRNA against Trim21 (siTrim21), and cell viability was assessed 1 and 6 h after S. Typhimurium infection by trypan blue staining. Cell viability is displayed relative to uninfected siCtrl BMDMs. (B) Expression levels of TRIM21, p62, and NRF2 in total cell lysates of S. Typhimurium–infected WT BMDMs were analyzed by immunoblot after transfection with either control (siCtrl) or Trim21-specific siRNA (siTrim21). β-Actin served as a loading control. (CE) Densitometric quantification of (C) TRIM21, (D) p62, and (E) NRF2 signal intensities from five independent immunoblots normalized to the corresponding β-actin bands. (F and G) WT BMDMs transfected with control (siCtrl) or Trim21-specific siRNA (siTrim21) were infected with S. Typhimurium for 1 and 6 h, respectively, and relative mRNA expression levels of (F) Hmox-1 and (G) Gclc were determined by quantitative real-time PCR (qRT-PCR). Values were normalized to expression levels of uninfected siCtrl BMDMs. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. Hmox-1, heme oxygenase 1;. Gclc, Glutamate cysteine ligase, catalytic subunit; Hmox-1, heme oxygenase 1; ST, S. Typhimurium; UI, uninfected.

FIGURE 2.

TRIM21 promotes cell death. (A) WT BMDMs were transfected with nontargeting siRNA (siCtrl) or siRNA against Trim21 (siTrim21), and cell viability was assessed 1 and 6 h after S. Typhimurium infection by trypan blue staining. Cell viability is displayed relative to uninfected siCtrl BMDMs. (B) Expression levels of TRIM21, p62, and NRF2 in total cell lysates of S. Typhimurium–infected WT BMDMs were analyzed by immunoblot after transfection with either control (siCtrl) or Trim21-specific siRNA (siTrim21). β-Actin served as a loading control. (CE) Densitometric quantification of (C) TRIM21, (D) p62, and (E) NRF2 signal intensities from five independent immunoblots normalized to the corresponding β-actin bands. (F and G) WT BMDMs transfected with control (siCtrl) or Trim21-specific siRNA (siTrim21) were infected with S. Typhimurium for 1 and 6 h, respectively, and relative mRNA expression levels of (F) Hmox-1 and (G) Gclc were determined by quantitative real-time PCR (qRT-PCR). Values were normalized to expression levels of uninfected siCtrl BMDMs. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. Hmox-1, heme oxygenase 1;. Gclc, Glutamate cysteine ligase, catalytic subunit; Hmox-1, heme oxygenase 1; ST, S. Typhimurium; UI, uninfected.

Close modal

Having demonstrated that TRIM21 promotes cell death, we next sought to identify the pathways that regulate posttranscriptional TRIM21 expression during S. Typhimurium infection in more detail. Although our results indicate that Trim21 mRNA expression in WT BMDMs was highly upregulated 1 h after S. Typhimurium infection (Fig. 1A), the corresponding protein levels at 1 h p.i. were markedly reduced compared with uninfected controls (Fig. 1B, 1C). We therefore assumed that TRIM21 was transiently degraded following S. Typhimurium infection.

Other proteins of the TRIM family, such as TRIM56 and TRIM65, are reported to be ubiquitinated by the SPI-1–encoded bacterial ubiquitin ligase SopA, which resulted in their proteasomal degradation (29, 30). To investigate whether transient degradation of TRIM21 is mediated by SopA or other SPI-1–encoded proteins, we infected WT BMDMs with either sopA- or SPI-1–deficient S. Typhimurium mutant strains (ΔsopA or ΔSPI-1) and analyzed TRIM21 expression by immunoblot. As shown in Fig. 3A, 3B, infection of WT BMDMs with S. Typhimurium ΔsopA or ΔSPI-1 mutant strains did not prevent degradation of TRIM21 at 1 h p.i. Because it has been suggested that only Ab-decorated pathogens can interact with TRIM21 because of its function as an intracellular Fc receptor (31), we incubated S. Typhimurium with murine serum prior to infection. As shown in Supplemental Fig. 2A, 2B, prior serum opsonization did not alter the expression pattern of TRIM21 during S. Typhimurium infection.

FIGURE 3.

TRIM21 is transiently degraded following TLR4 activation. (A) WT BMDMs were either infected with S. Typhimurium WT or ΔSPI-1 or ΔsopA mutant strains. Alternatively, noninfected WT BMDMs were treated with 200 ng/ml LPS. At the indicated time points, BMDMs were lysed and subjected to immunoblot analysis to determine TRIM21 expression. β-Actin served as a loading control. (B) TRIM21 band intensities were quantified densitometrically from three independent immunoblots relative to the corresponding β-actin signals. (C) Cell lysates from MyD88−/− and TRIF−/− BMDMs infected with S. Typhimurium for 1, 4, and 6 h, respectively, were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin was used as a loading control. (D) Densitometric quantification of TRIM21 signal intensities from three independent immunoblots normalized to the corresponding β-actin bands. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

FIGURE 3.

TRIM21 is transiently degraded following TLR4 activation. (A) WT BMDMs were either infected with S. Typhimurium WT or ΔSPI-1 or ΔsopA mutant strains. Alternatively, noninfected WT BMDMs were treated with 200 ng/ml LPS. At the indicated time points, BMDMs were lysed and subjected to immunoblot analysis to determine TRIM21 expression. β-Actin served as a loading control. (B) TRIM21 band intensities were quantified densitometrically from three independent immunoblots relative to the corresponding β-actin signals. (C) Cell lysates from MyD88−/− and TRIF−/− BMDMs infected with S. Typhimurium for 1, 4, and 6 h, respectively, were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin was used as a loading control. (D) Densitometric quantification of TRIM21 signal intensities from three independent immunoblots normalized to the corresponding β-actin bands. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

Close modal

To further exclude that TRIM21 degradation was mediated by other bacterial virulence proteins, we treated WT BMDMs with the TLR4 agonist LPS, which is a part of the outer membrane of Gram-negative bacteria, and determined TRIM21 expression by immunoblot. Similarly, engagement of TLR4 resulted in transient degradation of TRIM21 after 1 h of infection (Fig. 3A, 3B). We further confirmed TLR4-dependent degradation of TRIM21 by infecting WT BMDMs and BMDMs that lack the TLR4 adaptor proteins MyD88 or TRIF, respectively. As shown in Fig. 3C, loss of MyD88 did not affect TRIM21 degradation at 1 h p.i., whereas TRIM21 levels were reduced throughout infection in the absence of TRIF. From these results, we conclude that S. Typhimurium infection of macrophages results in the transient degradation of TRIM21, which occurs independent of bacterial virulence factors but dependent on bacterial activation of TLR4–TRIF axis.

Having shown that TRIM21 was transiently degraded following TLR4 activation, we next sought to identify the pathway that mediates TRIM21 degradation. Because E3 ubiquitin ligase regulation is thought to be governed by self-ubiquitination and subsequent proteasomal degradation (32), we first analyzed whether TRIM21 was degraded via the proteasomal pathway. We therefore treated WT BMDMs with the proteasomal inhibitor MG132 prior to S. Typhimurium infection and analyzed TRIM21 expression by immunoblot. As shown in Fig. 4A, 4B, MG132 treatment did not prevent TRIM21 degradation at 1 and 4 h of infection, rather it led to a further decrease in TRIM21 levels following infection, suggesting another degradation pathway is involved.

FIGURE 4.

TRIM21 is degraded via the lysosomal pathway. (A) WT BMDMs were infected with S. Typhimurium in the absence or presence of MG132 (10 μM) or Concanamycin A (ConcA; 100 nM). MG132 inhibits the degradation of ubiquitinated proteins by the 26S proteasome, whereas ConcA inhibits ATP-dependent acidification of the lysosome. At the indicated timepoints, TRIM21 expression in total cell lysates was determined by immunoblot, and β-actin was used as a loading control. (B) TRIM21 signal intensities from three independent immunoblots were assessed by densitometric quantification relative to the corresponding β-actin bands. (C) WT BMDMs were infected with S. Typhimurium for 1 and 6 h, respectively, and recruitment of TRIM21 (green) to the lysosomal compartment was analyzed by immunofluorescence microscopy using LysoTracker (red) staining. LysoTracker is a red-fluorescent dye that is sequestered in acidic organelles, predominantly lysosomes. Intracellular bacteria were stained with an Ab against LPS (blue). Scale bar, 10 μm. (D) Pearson correlation coefficient was determined to quantify the degree of colocalization between TRIM21 and LysoTracker. Fifty cells from three independent experiments were counted per time point. (E) S. Typhimurium–infected WT BMDMs were treated with either MG132 or ConcA as described in (A). At the indicated time points, cell lysates were subjected to immunoblot analysis and TRIM21, and LC3-I and LC3-II expression was determined using specific Abs. β-Actin served as loading control throughout. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected; UT, untreated.

FIGURE 4.

TRIM21 is degraded via the lysosomal pathway. (A) WT BMDMs were infected with S. Typhimurium in the absence or presence of MG132 (10 μM) or Concanamycin A (ConcA; 100 nM). MG132 inhibits the degradation of ubiquitinated proteins by the 26S proteasome, whereas ConcA inhibits ATP-dependent acidification of the lysosome. At the indicated timepoints, TRIM21 expression in total cell lysates was determined by immunoblot, and β-actin was used as a loading control. (B) TRIM21 signal intensities from three independent immunoblots were assessed by densitometric quantification relative to the corresponding β-actin bands. (C) WT BMDMs were infected with S. Typhimurium for 1 and 6 h, respectively, and recruitment of TRIM21 (green) to the lysosomal compartment was analyzed by immunofluorescence microscopy using LysoTracker (red) staining. LysoTracker is a red-fluorescent dye that is sequestered in acidic organelles, predominantly lysosomes. Intracellular bacteria were stained with an Ab against LPS (blue). Scale bar, 10 μm. (D) Pearson correlation coefficient was determined to quantify the degree of colocalization between TRIM21 and LysoTracker. Fifty cells from three independent experiments were counted per time point. (E) S. Typhimurium–infected WT BMDMs were treated with either MG132 or ConcA as described in (A). At the indicated time points, cell lysates were subjected to immunoblot analysis and TRIM21, and LC3-I and LC3-II expression was determined using specific Abs. β-Actin served as loading control throughout. **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected; UT, untreated.

Close modal

Apart from the ubiquitin proteasome system, proteins can also be degraded via the lysosomal pathway (33). To investigate whether S. Typhimurium infection induces degradation of TRIM21 through the lysosomal pathway, we infected WT BMDMs with S. Typhimurium in the presence of the lysosomal inhibitor Concanamycin A and determined TRIM21 levels by immunoblot. Compared with untreated samples, Concanamycin A attenuated TRIM21 degradation at 1 and 4 h p.i. (Fig. 4A, 4B). We further confirmed that TRIM21 was degraded through the lysosomal pathway by LysoTracker staining, a fluorescent dye that accumulates in the lysosomal compartment. In line with our immunoblot data, TRIM21 strongly colocalized with LysoTracker 1 and 6 h p.i. (Fig. 4C, 4D), indicating that it was recruited to lysosomes.

Autophagy is the catabolic process that culminates in lysosomal degradation in eukaryotes. Upon canonical autophagy induction, damaged cellular organelles and proteins are engulfed by the autophagosome, which subsequently fuses with the lysosome, leading to the enzymatic degradation of the cargo (4, 5). We and others have previously reported that S. Typhimurium interferes with autophagy, thereby altering intracellular protein levels to impair the host’s antimicrobial response (3, 26, 34). To analyze whether TRIM21 degradation correlates with autophagy induction, we determined conversion of LC3-I into lipidated LC3-II, which is recruited to and degraded along with the autophagosome. As shown in Fig. 4E, S. Typhimurium infection transiently triggered LC3-I–LC3-II conversion at 1 h p.i., indicative of autophagy induction, whereas LC3-I–LC3-II conversion was reduced at 4 and 6 h p.i., respectively, indicative of autophagy inhibition. Interestingly, MG132 treatment induced LC3-I–LC3-II conversion, suggesting that autophagic pathways may be activated to compensate for proteasome inhibition. By contrast, autophagic degradation was strongly inhibited upon lysosomal inhibition with Concanamycin A resulting in LC3-II accumulation (Fig. 4E). Accordingly, TRIM21 levels were significantly elevated upon Concanamycin A treatment (Fig. 4A, 4B), indicating that TRIM21 expression levels inversely correlated with autophagic activity.

To further investigate whether TRIM21 is targeted by canonical autophagy, we infected WT and Atg7−/− BMDMs with S. Typhimurium and analyzed TRIM21 expression by immunoblot. ATG7 facilitates conjugation of phosphatidylethanolamine to LC3, which is an upstream event required for autophagosome formation and a characteristic feature of canonical autophagy (35). Compared with WT BMDMs, loss of ATG7 had no significant effect on TRIM21 expression following S. Typhimurium infection (Supplemental Fig. 3A, 3B). Together, these findings indicate that TRIM21 is transiently degraded through the lysosomal pathway independent of canonical autophagy.

In addition to canonical autophagy, CMA represents an alternative pathway that culminates in lysosomal degradation. CMA is a selective form of autophagy by which substrate proteins are bound to HSC70, a member of the family of heat shock protein 70 (HSP70), and delivered to the lysosomal membrane, where they are taken up in an LAMP2A-dependent manner (36).

Proteins recognized by HSC70 share conserved sequence homologies known as KFERQ-like motifs (37). To investigate whether TRIM21 is a potential substrate of CMA, we first analyzed whether TRIM21 contains any KFERQ-like motifs using the KFERQ finder software (38). We identified at least two canonical KFERQ-like motifs within the protein sequence of TRIM21 (Fig. 5A), indicating that TRIM21 potentially interacts with HSC70. In line with this notion, TRIM21 colocalized with HSC70 after 1 and 6 h of S. Typhimurium infection, as assessed by immunofluorescence staining (Fig. 5B, 5C). Furthermore, TRIM21 was delivered to LAMP2A-positive lysosomes at 1 and 6 h p.i. with S. Typhimurium (Fig. 5D, 5E), which is a characteristic feature of CMA (39). These results therefore suggest that during S. Typhimurium infection, TRIM21 is transiently degraded via CMA.

FIGURE 5.

TRIM21 is degraded by CMA. (A) KFERQ-like motifs within the protein sequence of TRIM21 (UniProt identifier [ID] Q62191) were identified using the KFERQ finder software v0.8 (38). The presence of a KFERQ-like motif indicates a putative HSC70 binding site, which is required for recognition of CMA target proteins. KFERQ-like motifs were underlined, and canonical motifs were additionally highlighted in gray. (B) Colocalization of TRIM21 (red) with the chaperone HSC70 (green) in S. Typhimurium–infected WT BMDMs was assessed by immunofluorescence microscopy. Scale bar, 10 μm. Each inset represents a magnified image. (C) Pearson correlation coefficient was determined to quantify the degree of colocalization between TRIM21 and HSC70. Fifty cells from three independent experiments were counted per time point. (D) Colocalization of TRIM21 (red) with the lysosomal receptor LAMP2A (green) in S. Typhimurium–infected WT BMDMs was analyzed by immunofluorescence microscopy. Scale bar, 10 μm. Each inset represents a magnified image. (E) Pearson correlation coefficient was calculated to quantify the degree of colocalization between TRIM21 and LAMP2A. Fifty cells from three independent experiments were counted per time point. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

FIGURE 5.

TRIM21 is degraded by CMA. (A) KFERQ-like motifs within the protein sequence of TRIM21 (UniProt identifier [ID] Q62191) were identified using the KFERQ finder software v0.8 (38). The presence of a KFERQ-like motif indicates a putative HSC70 binding site, which is required for recognition of CMA target proteins. KFERQ-like motifs were underlined, and canonical motifs were additionally highlighted in gray. (B) Colocalization of TRIM21 (red) with the chaperone HSC70 (green) in S. Typhimurium–infected WT BMDMs was assessed by immunofluorescence microscopy. Scale bar, 10 μm. Each inset represents a magnified image. (C) Pearson correlation coefficient was determined to quantify the degree of colocalization between TRIM21 and HSC70. Fifty cells from three independent experiments were counted per time point. (D) Colocalization of TRIM21 (red) with the lysosomal receptor LAMP2A (green) in S. Typhimurium–infected WT BMDMs was analyzed by immunofluorescence microscopy. Scale bar, 10 μm. Each inset represents a magnified image. (E) Pearson correlation coefficient was calculated to quantify the degree of colocalization between TRIM21 and LAMP2A. Fifty cells from three independent experiments were counted per time point. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. ST, S. Typhimurium; UI, uninfected.

Close modal

The mTORC2/Akt pathway has recently been recognized as an important regulator of CMA and lysosomal degradation. Activation of mTORC2 mediates phosphorylation of Akt at serine 473 (S473), which exerts an inhibitory effect on CMA and lysosomal degradation (7).

We therefore investigated whether S. Typhimurium infection of macrophages interferes with mTORC2/Akt signaling to regulate lysosomal degradation of TRIM21. Our results indicated that S. Typhimurium infection of WT BMDMs led to time-dependent activation of mTORC2, as assessed by expression of Rictor, an essential component of mTORC2, and phosphorylation of Akt S473 (Fig. 6A–C). Importantly, activation of mTORC2/Akt signaling at 6 h p.i. positively correlated with TRIM21 expression (Fig. 6A, 6D).

FIGURE 6.

TRIM21 levels are regulated by mTORC2/Akt. (A) WT BMDMs were infected with S. Typhimurium for the indicated time in the presence or absence of Akt inhibitor VIII (10 μM). Total cell lysates were analyzed by immunoblot for expression of TRIM21, Rictor, and Akt phosphorylated at serine 473 (p-Akt S473), which is a downstream target of mTORC2. β-Actin was used as a loading control. (BD) Signal intensities of (B) Rictor, (C) p-Akt S473, and (D) TRIM21 were determined densitometrically from three independent immunoblots relative to the signals of the corresponding β-actin bands. (E) S. Typhimurium–infected WT BMDMs were either treated with Torin1 (10 μM) or rapamycin (0.5 μM) for the indicated time. Torin1 inhibits both mTORC1 and mTORC2 activity, whereas rapamycin specifically inhibits mTORC1. Total cell lysates were subjected to immunoblot and analyzed for TRIM21 expression. β-Actin served a loading control. (F) Signal intensities of TRIM21 were determined densitometrically from three independent immunoblots relative to the signals of the corresponding β-actin bands. (G) WT BMDMs were transfected with either control siRNA (siCtrl) or siRNA specific against Rictor (siRictor). After S. Typhimurium infection for 1 and 6 h, BMDMs were lysed, and expression levels of Rictor and p-Akt S473 were determined by immunoblot. β-Actin served as a loading control. (H and I) Expression levels of (H) Rictor and (I) p-Akt S473 were assessed densitometrically relative to the corresponding β-actin signals from three independent immunoblots. (J) WT BMDMs were transfected and infected with S. Typhimurium as described in (G). At the indicated time points, total cell lysates were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin served as a loading control. (K) TRIM21 signal intensities were quantified densitometrically relative to the corresponding β-actin signals from three independent immunoblots. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. Rapa, rapamycin; ST, S. Typhimurium; UI, uninfected; UT, untreated.

FIGURE 6.

TRIM21 levels are regulated by mTORC2/Akt. (A) WT BMDMs were infected with S. Typhimurium for the indicated time in the presence or absence of Akt inhibitor VIII (10 μM). Total cell lysates were analyzed by immunoblot for expression of TRIM21, Rictor, and Akt phosphorylated at serine 473 (p-Akt S473), which is a downstream target of mTORC2. β-Actin was used as a loading control. (BD) Signal intensities of (B) Rictor, (C) p-Akt S473, and (D) TRIM21 were determined densitometrically from three independent immunoblots relative to the signals of the corresponding β-actin bands. (E) S. Typhimurium–infected WT BMDMs were either treated with Torin1 (10 μM) or rapamycin (0.5 μM) for the indicated time. Torin1 inhibits both mTORC1 and mTORC2 activity, whereas rapamycin specifically inhibits mTORC1. Total cell lysates were subjected to immunoblot and analyzed for TRIM21 expression. β-Actin served a loading control. (F) Signal intensities of TRIM21 were determined densitometrically from three independent immunoblots relative to the signals of the corresponding β-actin bands. (G) WT BMDMs were transfected with either control siRNA (siCtrl) or siRNA specific against Rictor (siRictor). After S. Typhimurium infection for 1 and 6 h, BMDMs were lysed, and expression levels of Rictor and p-Akt S473 were determined by immunoblot. β-Actin served as a loading control. (H and I) Expression levels of (H) Rictor and (I) p-Akt S473 were assessed densitometrically relative to the corresponding β-actin signals from three independent immunoblots. (J) WT BMDMs were transfected and infected with S. Typhimurium as described in (G). At the indicated time points, total cell lysates were subjected to immunoblot analysis, and TRIM21 expression was determined using specific Abs. β-Actin served as a loading control. (K) TRIM21 signal intensities were quantified densitometrically relative to the corresponding β-actin signals from three independent immunoblots. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. Rapa, rapamycin; ST, S. Typhimurium; UI, uninfected; UT, untreated.

Close modal

To further confirm that TRIM21 levels were regulated by mTORC2/Akt, we treated WT BMDMs with an Akt inhibitor and determined TRIM21 expression. Compared with untreated macrophages, the amount of TRIM21 at 6 h of S. Typhimurium infection was significantly decreased when Akt S473 phosphorylation was inhibited (Fig. 6A, 6D). Likewise, treating WT BMDMs with Torin1 (which inhibits both mTORC1 and mTORC2) resulted in reduced TRIM21 levels at 6 h p.i. compared with untreated controls (Fig. 6E, 6F). By contrast, rapamycin treatment (which specifically inhibits mTORC1) had no significant effect on TRIM21 expression at 6 h p.i. compared with untreated controls (Fig. 6E, 6F).

Next, we transfected WT BMDMs with siRNA against Rictor (Fig. 6G, 6H). As expected, genetic silencing of Rictor resulted in reduced phosphorylation of Akt S473 compared with WT BMDMs transfected with control siRNA (Fig. 6G, 6I). Notably, TRIM21 expression in S. Typhimurium–infected WT BMDMs was significantly reduced following knockdown of Rictor at 6 h p.i. (Fig. 6J, 6K).

In summary, we conclude that S. Typhimurium induces mTORC2/Akt signaling to enhance TRIM21 expression by attenuating its lysosomal degradation through CMA (Fig. 7).

FIGURE 7.

S. Typhimurium induces mTORC2/Akt signaling to regulate TRIM21 levels through CMA. In S. Typhimurium–infected macrophages, the E3 ubiquitin ligase TRIM21 is induced following activation of the IFN-I receptor. Because TRIM21 promotes cell death, its expression levels are tightly controlled posttranscriptionally by targeting TRIM21 for CMA through LAMP2A-positive lysosomes. S. Typhimurium subsequently induces mTORC2/Akt signaling to increase TRIM21 levels by impairing its CMA-mediated degradation, which contributes to cell death.

FIGURE 7.

S. Typhimurium induces mTORC2/Akt signaling to regulate TRIM21 levels through CMA. In S. Typhimurium–infected macrophages, the E3 ubiquitin ligase TRIM21 is induced following activation of the IFN-I receptor. Because TRIM21 promotes cell death, its expression levels are tightly controlled posttranscriptionally by targeting TRIM21 for CMA through LAMP2A-positive lysosomes. S. Typhimurium subsequently induces mTORC2/Akt signaling to increase TRIM21 levels by impairing its CMA-mediated degradation, which contributes to cell death.

Close modal

TRIM21 has been identified as an important regulator of innate immune responses because of its ability to ubiquitinate proteins involved in the antimicrobial response as well as interfering with the IFN-I system (14). Although many studies have highlighted the importance of TRIM21 for antiviral immunity, relatively little is known on its role during bacterial infection. We report in this study that S. Typhimurium infection of macrophages induces the transcription of Trim21 in an IFN-I–dependent manner. Elevated TRIM21 levels sensitize S. Typhimurium–infected macrophages to cell death correlating with reduced p62/NRF2-dependent antioxidative stress responses. Because TRIM21 negatively regulates infection outcome, its posttranscriptional expression is tightly controlled by targeting TRIM21 for CMA. However, S. Typhimurium induces mTORC2/Akt signaling to enhance TRIM21 levels through impairing its CMA-mediated degradation (Fig. 7).

In recent years, phylogenetic studies revealed that the TRIM protein family is closely linked and may have coevolved with the IFN-I system (14). In line with this notion, we found that Trim21 was upregulated following engagement of the IFN-I receptor in S. Typhimurium–infected macrophages. By contrast, TRIM21 was barely detectable in macrophages that lacked the IFN-I receptor. These results are in accordance with previous studies demonstrating that many TRIMs, including TRIM21, are inducible by IFN signaling (40, 41). Interestingly, TRIM21 has also been reported to act upstream of IFN-I induction as it mediates IRF3 degradation, hence limiting IRF3-dependent IFN-I production (20).

Although our results indicate that Trim21 is transcriptionally upregulated in response to S. Typhimurium infection, TRIM21 expression is mainly regulated at the posttranscriptional level. In fact, S. Typhimurium infection resulted in a transient degradation of TRIM21, which was driven by bacterial engagement of TLR4. To date, most studies assessing TRIM function in innate immunity have investigated the effect of TRIM-mediated ubiquitination on target protein function. It should be noted, however, that several TRIM proteins undergo posttranslational modification themselves. For instance, the linear ubiquitination assembly complex (Lubac) mediates K48-linked ubiquitination of TRIM25, resulting in its proteasomal degradation (42). Similarly, other TRIM proteins are degraded through the proteasome system following autoubiquitination or ubiquitination by yet-unknown ubiquitin ligases (41, 43). Interestingly, many Gram-negative bacteria, including S. Typhimurium, counteract innate immune responses by interfering with the eukaryotic ubiquitin system (17, 44). During S. Typhimurium infection, TRIM56 and TRIM65 have recently been reported to undergo proteasomal degradation following ubiquitination by the SPI-1–encoded effector protein SopA, which acts as a bacterial E3 ubiquitin ligase (30). Despite close sequence similarities among TRIM proteins, SopA interacted exclusively with the RING domains of TRIM56 and TRIM65 (30). In line with this observation, we report in this study that neither SopA nor other SPI-1–encoded effector proteins were involved in the degradation of TRIM21.

In contrast to previous studies on other members of the TRIM family, we report in this study that degradation of TRIM21 occurs independent of the proteasomal system. In fact, pharmacological inhibition of the proteasome even decreased TRIM21 levels throughout infection, presumably by upregulating autophagic pathways, including CMA, as a compensatory mechanism to maintain protein homeostasis (33, 45). In line with this notion, we demonstrate that, unlike other TRIM proteins, TRIM21 is transiently degraded through the lysosomal pathway. Lysosomal degradation of TRIM21 occurred independent of canonical autophagy because loss of the essential autophagy protein ATG7 and pharmacological inhibition of mTORC1, a master regulator of autophagy, did not significantly alter TRIM21 levels. Instead, we found that TRIM21 was delivered to the lysosomal compartment by CMA. CMA has previously been recognized as a selective lysosomal degradation process that is induced in response to various cellular stressors, such as starvation or oxidative stress (46, 47). Proteins degraded by CMA contain a KFERQ-like motif in their amino acid sequence that allows binding of the cytosolic chaperone HSC70, which selectively delivers substrates to LAMP2A on lysosomal membranes (36, 48). Previous studies showed that the presence of a KFERQ-like motif in a protein sequence is necessary and sufficient for its targeting and degradation via CMA (37, 38). Accordingly, we found that TRIM21 contained at least two KFERQ-like motifs and colocalized with the chaperone HSC70 as well as with LAMP2A-positive lysosomes, demonstrating that it was targeted for CMA during S. Typhimurium infection. Although HSC70 is recruited to LAMP2A monomers and both proteins act at the same location of the lysosome, our data show that TRIM21 colocalized less with HSC70 than with LAMP2A. This difference could be due to transient TRIM21/HSC70 colocalization. Indeed, recent work demonstrated that cytosolic HSC70 was released from LAMP2A before the fully functional LAMP2A translocation complex was assembled (49, 50), indicating that HSC70, unlike LAMP2A, is only transiently present at the lysosomal membrane. Recently, we and others reported that S. Typhimurium infection or cellular stress activates mTORC2/Akt signaling, which disrupts lysosomal activity (7, 51). Importantly, Akt activation had no effect on HSC70-mediated recruitment of CMA substrates to the lysosomal membrane, whereas it impaired LAMP2A-mediated uptake and subsequent lysosomal degradation (7). Similarly, our results demonstrate that S. Typhimurium–induced phosphorylation of Akt at Ser473 enhanced TRIM21 levels through impairment of CMA. Although mTORC2/Akt signaling had no significant effect on HSC70-mediated recruitment of TRIM21 to LAMP2A-positive lysosomes, it impaired the lysosomal uptake and degradation of TRIM21, leading to accumulation of TRIM21 on lysosomal membranes. Conversely, pharmacological blockade or genetic silencing of mTORC2/Akt resulted in decreased TRIM21 levels, consistent with the reported inhibitory effect of mTORC2/Akt on CMA and lysosomal degradation (7, 51).

Dependent on the pathogen and cell type, TRIM21 can have both positive and negative effects on infection outcome (15, 21). Our results indicate that high levels of TRIM21 sensitized S. Typhimurium–infected macrophages to cell death. Mechanistically, we found that TRIM21 impaired the transcription of NRF2-dependent cytoprotective genes, resulting in a reduced antioxidative stress response to S. Typhimurium infection. We have previously reported that IFN-I signaling during S. Typhimurium infection was detrimental for the host, as it induced macrophage death through reducing p62 levels and interfering with NRF2-dependent stress responses (26). Under quiescent conditions, the transcription factor NRF2 was sequestered in the cytosol by interacting with KEAP1, which resulted in proteasomal degradation of NRF2. Oxidative stress, however, caused a conformational change of KEAP1 allowing NRF2 to stabilize and translocate to the nucleus, where it induced the transcription of several cytoprotective genes (52). Furthermore, high levels of p62 promoted nuclear translocation of NRF2, as p62 competitively interacted with the NRF2-binding site of KEAP1 (53). Importantly, TRIM21 has previously been shown to impair NRF2-dependent antioxidative responses through ubiquitination of p62 at K7 via K63 linkage, which prevented p62–KEAP1 interaction and, hence, promoted binding of NRF2 to KEAP1 (16). In line with these findings, we demonstrate in this study that TRIM21 negatively regulated NRF2-dependent antioxidative pathways through reducing p62 and NRF2 levels. High amounts of TRIM21 correlated with reduced transcription of NRF2-dependent genes and enhanced macrophage death following S. Typhimurium infection.

In summary, to our knowledge, our results provide new insight into the dynamic interplay of S. Typhimurium with the host’s E3 ubiquitin ligase TRIM21, which is an important component of the IFN-I–induced immune response. To our knowledge, this is the first study identifying TRIM21 as a substrate of CMA and lysosomal degradation. Given the importance of lysosomal degradation processes for antibacterial host defense, pharmacological interference with CMA might be a promising approach to modulate innate immune responses against bacterial infections.

We thank the Robinson’s laboratory members for helpful discussion and Monika Keiten Schmitz for technical support. Ifnar1−/− mice were kindly provided by Zeinab Abdullah (Institute of Experimental Immunology, University of Bonn, Germany) and Atg7−/−, MyD88−/−, and TRIF−/−mice were obtained from Michael Schramm (Institute for Medical Microbiology, Immunology and Hygiene, University of Cologne, Germany). The S. Typhimurium ΔSPI-1 and ΔsopA mutant strains were a kind gift from Ivan Dikic (Institute for Biochemistry II, Goethe University Frankfurt, Germany).

This work was supported by research grants from the German Center for Infection Research (to N.J.H. and J.F.), the Center for Molecular Medicine Cologne, University of Cologne, Cologne, Germany (to J.F., D.H., and J.R.), the Cologne Fortune Program of the University of Cologne (to N.J.H., J.F., and D.H.), and the Cluster of Excellence in Cellular Stress Responses in Aging-Associated Diseases, University of Cologne (funded by the Deutsche Forschungsgemeinschaft within the Excellence Initiative of the German federal and state governments) (to N.R.). This work was also supported by Deutsche Forschungsgemeinschaft Grants SFB670 (to N.R. and M.K.) and FOR2240 (to D.H.).

The online version of this article contains supplemental material.

Abbreviations used in this article:

autophagy

macroautophagy

BHI

brain–heart infusion

BMDM

bone marrow–derived macrophage

CMA

chaperone-mediated autophagy

IFN-I

type I IFN

IRF

IFN regulatory factor

KEAP1

Kelch-like ECH-associated protein 1

LAMP2A

lysosome-associated membrane protein type 2A

mTORC

mammalian target of rapamycin complex

NRF2

NF erythroid 2–related factor 2

p62

sequestome 1

p.i.

postinfection

siCtrl

nontargeting small interfering RNA

siTrim21

small interfering RNA against Trim21

S. Typhimurium

Salmonella enterica serovar Typhimurium

siRNA

small interfering RNA

SPI

Salmonella pathogenicity island

TRIM

tripartite motif

WT

wild-type.

1
Broz
,
P.
,
M. B.
Ohlson
,
D. M.
Monack
.
2012
.
Innate immune response to Salmonella typhimurium, a model enteric pathogen.
Gut Microbes
3
:
62
70
.
2
Hensel
,
M.
,
J. E.
Shea
,
C.
Gleeson
,
M. D.
Jones
,
E.
Dalton
,
D. W.
Holden
.
1995
.
Simultaneous identification of bacterial virulence genes by negative selection.
Science
269
:
400
403
.
3
Ganesan
,
R.
,
N. J.
Hos
,
S.
Gutierrez
,
J.
Fischer
,
J. M.
Stepek
,
E.
Daglidu
,
M.
Krönke
,
N.
Robinson
.
2017
.
Salmonella Typhimurium disrupts Sirt1/AMPK checkpoint control of mTOR to impair autophagy.
PLoS Pathog.
13
: e1006227.
4
Deretic
,
V.
,
T.
Saitoh
,
S.
Akira
.
2013
.
Autophagy in infection, inflammation and immunity.
Nat. Rev. Immunol.
13
:
722
737
.
5
Mizushima
,
N.
,
B.
Levine
,
A. M.
Cuervo
,
D. J.
Klionsky
.
2008
.
Autophagy fights disease through cellular self-digestion.
Nature
451
:
1069
1075
.
6
Kim
,
J.
,
M.
Kundu
,
B.
Viollet
,
K. L.
Guan
.
2011
.
AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1.
Nat. Cell Biol.
13
:
132
141
.
7
Arias
,
E.
,
H.
Koga
,
A.
Diaz
,
E.
Mocholi
,
B.
Patel
,
A. M.
Cuervo
.
2015
.
Lysosomal mTORC2/PHLPP1/Akt regulate chaperone-mediated autophagy.
Mol. Cell
59
:
270
284
.
8
Perrin
,
A. J.
,
X.
Jiang
,
C. L.
Birmingham
,
N. S.
So
,
J. H.
Brumell
.
2004
.
Recognition of bacteria in the cytosol of mammalian cells by the ubiquitin system.
Curr. Biol.
14
:
806
811
.
9
Matsumoto
,
G.
,
K.
Wada
,
M.
Okuno
,
M.
Kurosawa
,
N.
Nukina
.
2011
.
Serine 403 phosphorylation of p62/SQSTM1 regulates selective autophagic clearance of ubiquitinated proteins.
Mol. Cell
44
:
279
289
.
10
Liu
,
W. J.
,
L.
Ye
,
W. F.
Huang
,
L. J.
Guo
,
Z. G.
Xu
,
H. L.
Wu
,
C.
Yang
,
H. F.
Liu
.
2016
.
p62 links the autophagy pathway and the ubiqutin-proteasome system upon ubiquitinated protein degradation.
Cell. Mol. Biol. Lett.
21
:
29
.
11
Ben-Chetrit
,
E.
,
E. K.
Chan
,
K. F.
Sullivan
,
E. M.
Tan
.
1988
.
A 52-kD protein is a novel component of the SS-A/Ro antigenic particle.
J. Exp. Med.
167
:
1560
1571
.
12
Ben-Chetrit
,
E.
,
R. I.
Fox
,
E. M.
Tan
.
1990
.
Dissociation of immune responses to the SS-A (Ro) 52-kd and 60-kd polypeptides in systemic lupus erythematosus and Sjögren’s syndrome.
Arthritis Rheum.
33
:
349
355
.
13
Wada
,
K.
,
T.
Kamitani
.
2006
.
Autoantigen Ro52 is an E3 ubiquitin ligase.
Biochem. Biophys. Res. Commun.
339
:
415
421
.
14
Versteeg
,
G. A.
,
S.
Benke
,
A.
García-Sastre
,
R.
Rajsbaum
.
2014
.
InTRIMsic immunity: positive and negative regulation of immune signaling by tripartite motif proteins.
Cytokine Growth Factor Rev.
25
:
563
576
.
15
Zhang
,
Z.
,
M.
Bao
,
N.
Lu
,
L.
Weng
,
B.
Yuan
,
Y. J.
Liu
.
2013
.
The E3 ubiquitin ligase TRIM21 negatively regulates the innate immune response to intracellular double-stranded DNA.
Nat. Immunol.
14
:
172
178
.
16
Pan
,
J. A.
,
Y.
Sun
,
Y. P.
Jiang
,
A. J.
Bott
,
N.
Jaber
,
Z.
Dou
,
B.
Yang
,
J. S.
Chen
,
J. M.
Catanzaro
,
C.
Du
, et al
.
2016
.
TRIM21 ubiquitylates SQSTM1/p62 and suppresses protein sequestration to regulate redox homeostasis [Published erratum appears in 2016 Mol. Cell 62: 149–151.].
Mol. Cell
61
:
720
733
.
17
Herhaus
,
L.
,
I.
Dikic
.
2018
.
Regulation of Salmonella-host cell interactions via the ubiquitin system.
Int. J. Med. Microbiol.
308
:
176
184
.
18
Yau
,
R.
,
M.
Rape
.
2016
.
The increasing complexity of the ubiquitin code.
Nat. Cell Biol.
18
:
579
586
.
19
Yang
,
K.
,
H. X.
Shi
,
X. Y.
Liu
,
Y. F.
Shan
,
B.
Wei
,
S.
Chen
,
C.
Wang
.
2009
.
TRIM21 is essential to sustain IFN regulatory factor 3 activation during antiviral response.
J. Immunol.
182
:
3782
3792
.
20
Higgs
,
R.
,
J.
Ní Gabhann
,
N.
Ben Larbi
,
E. P.
Breen
,
K. A.
Fitzgerald
,
C. A.
Jefferies
.
2008
.
The E3 ubiquitin ligase Ro52 negatively regulates IFN-beta production post-pathogen recognition by polyubiquitin-mediated degradation of IRF3.
J. Immunol.
181
:
1780
1786
.
21
Vaysburd
,
M.
,
R. E.
Watkinson
,
H.
Cooper
,
M.
Reed
,
K.
O’Connell
,
J.
Smith
,
J.
Cruickshanks
,
L. C.
James
.
2013
.
Intracellular antibody receptor TRIM21 prevents fatal viral infection.
Proc. Natl. Acad. Sci. USA
110
:
12397
12401
.
22
Mancuso
,
G.
,
A.
Midiri
,
C.
Biondo
,
C.
Beninati
,
S.
Zummo
,
R.
Galbo
,
F.
Tomasello
,
M.
Gambuzza
,
G.
Macrì
,
A.
Ruggeri
, et al
.
2007
.
Type I IFN signaling is crucial for host resistance against different species of pathogenic bacteria.
J. Immunol.
178
:
3126
3133
.
23
Desvignes
,
L.
,
A. J.
Wolf
,
J. D.
Ernst
.
2012
.
Dynamic roles of type I and type II IFNs in early infection with Mycobacterium tuberculosis.
J. Immunol.
188
:
6205
6215
.
24
O’Connell
,
R. M.
,
S. K.
Saha
,
S. A.
Vaidya
,
K. W.
Bruhn
,
G. A.
Miranda
,
B.
Zarnegar
,
A. K.
Perry
,
B. O.
Nguyen
,
T. F.
Lane
,
T.
Taniguchi
, et al
.
2004
.
Type I interferon production enhances susceptibility to Listeria monocytogenes infection.
J. Exp. Med.
200
:
437
445
.
25
Robinson
,
N.
,
S.
McComb
,
R.
Mulligan
,
R.
Dudani
,
L.
Krishnan
,
S.
Sad
.
2012
.
Type I interferon induces necroptosis in macrophages during infection with Salmonella enterica serovar Typhimurium.
Nat. Immunol.
13
:
954
962
.
26
Hos
,
N. J.
,
R.
Ganesan
,
S.
Gutiérrez
,
D.
Hos
,
J.
Klimek
,
Z.
Abdullah
,
M.
Krönke
,
N.
Robinson
.
2017
.
Type I interferon enhances necroptosis of Salmonella Typhimurium-infected macrophages by impairing antioxidative stress responses.
J. Cell Biol.
216
:
4107
4121
.
27
Schneider
,
C. A.
,
W. S.
Rasband
,
K. W.
Eliceiri
.
2012
.
NIH Image to ImageJ: 25 years of image analysis.
Nat. Methods
9
:
671
675
.
28
Schindelin
,
J.
,
I.
Arganda-Carreras
,
E.
Frise
,
V.
Kaynig
,
M.
Longair
,
T.
Pietzsch
,
S.
Preibisch
,
C.
Rueden
,
S.
Saalfeld
,
B.
Schmid
, et al
.
2012
.
Fiji: an open-source platform for biological-image analysis.
Nat. Methods
9
:
676
682
.
29
Kamanova
,
J.
,
H.
Sun
,
M.
Lara-Tejero
,
J. E.
Galán
.
2016
.
The Salmonella effector protein SopA modulates innate immune responses by targeting TRIM E3 ligase family members.
PLoS Pathog.
12
: e1005552.
30
Fiskin
,
E.
,
S.
Bhogaraju
,
L.
Herhaus
,
S.
Kalayil
,
M.
Hahn
,
I.
Dikic
.
2017
.
Structural basis for the recognition and degradation of host TRIM proteins by Salmonella effector SopA.
Nat. Commun.
8
:
14004
.
31
McEwan
,
W. A.
,
J. C.
Tam
,
R. E.
Watkinson
,
S. R.
Bidgood
,
D. L.
Mallery
,
L. C.
James
.
2013
.
Intracellular antibody-bound pathogens stimulate immune signaling via the Fc receptor TRIM21.
Nat. Immunol.
14
:
327
336
.
32
Bhattacharyya
,
S.
,
H.
Yu
,
C.
Mim
,
A.
Matouschek
.
2014
.
Regulated protein turnover: snapshots of the proteasome in action.
Nat. Rev. Mol. Cell Biol.
15
:
122
133
.
33
Kocaturk
,
N. M.
,
D.
Gozuacik
.
2018
.
Crosstalk between mammalian autophagy and the ubiquitin-proteasome system.
Front. Cell Dev. Biol.
6
:
128
.
34
Owen
,
K. A.
,
C. B.
Meyer
,
A. H.
Bouton
,
J. E.
Casanova
.
2014
.
Activation of focal adhesion kinase by Salmonella suppresses autophagy via an Akt/mTOR signaling pathway and promotes bacterial survival in macrophages.
PLoS Pathog.
10
: e1004159.
35
Xiong
,
J.
2015
.
Atg7 in development and disease: panacea or Pandora’s Box?
Protein Cell
6
:
722
734
.
36
Cuervo
,
A. M.
,
J. F.
Dice
.
1996
.
A receptor for the selective uptake and degradation of proteins by lysosomes.
Science
273
:
501
503
.
37
Dice
,
J. F.
1982
.
Altered degradation of proteins microinjected into senescent human fibroblasts.
J. Biol. Chem.
257
:
14624
14627
.
38
Kirchner
,
P.
,
M.
Bourdenx
,
J.
Madrigal-Matute
,
S.
Tiano
,
A.
Diaz
,
B. A.
Bartholdy
,
B.
Will
,
A. M.
Cuervo
.
2019
.
Proteome-wide analysis of chaperone-mediated autophagy targeting motifs.
PLoS Biol.
17
: e3000301.
39
Tekirdag
,
K.
,
A. M.
Cuervo
.
2018
.
Chaperone-mediated autophagy and endosomal microautophagy: joint by a chaperone.
J. Biol. Chem.
293
:
5414
5424
.
40
Carthagena
,
L.
,
A.
Bergamaschi
,
J. M.
Luna
,
A.
David
,
P. D.
Uchil
,
F.
Margottin-Goguet
,
W.
Mothes
,
U.
Hazan
,
C.
Transy
,
G.
Pancino
,
S.
Nisole
.
2009
.
Human TRIM gene expression in response to interferons.
PLoS One
4
: e4894.
41
Versteeg
,
G. A.
,
R.
Rajsbaum
,
M. T.
Sánchez-Aparicio
,
A. M.
Maestre
,
J.
Valdiviezo
,
M.
Shi
,
K. S.
Inn
,
A.
Fernandez-Sesma
,
J.
Jung
,
A.
García-Sastre
.
2013
.
The E3-ligase TRIM family of proteins regulates signaling pathways triggered by innate immune pattern-recognition receptors.
Immunity
38
:
384
398
.
42
Inn
,
K. S.
,
M. U.
Gack
,
F.
Tokunaga
,
M.
Shi
,
L. Y.
Wong
,
K.
Iwai
,
J. U.
Jung
.
2011
.
Linear ubiquitin assembly complex negatively regulates RIG-I- and TRIM25-mediated type I interferon induction.
Mol. Cell
41
:
354
365
.
43
Mallery
,
D. L.
,
W. A.
McEwan
,
S. R.
Bidgood
,
G. J.
Towers
,
C. M.
Johnson
,
L. C.
James
.
2010
.
Antibodies mediate intracellular immunity through tripartite motif-containing 21 (TRIM21).
Proc. Natl. Acad. Sci. USA
107
:
19985
19990
.
44
Huibregtse
,
J.
,
J. R.
Rohde
.
2014
.
Hell’s BELs: bacterial E3 ligases that exploit the eukaryotic ubiquitin machinery.
PLoS Pathog.
10
: e1004255.
45
Koga
,
H.
,
M.
Martinez-Vicente
,
F.
Macian
,
V. V.
Verkhusha
,
A. M.
Cuervo
.
2011
.
A photoconvertible fluorescent reporter to track chaperone-mediated autophagy.
Nat. Commun.
2
:
386
.
46
Cuervo
,
A. M.
,
E.
Knecht
,
S. R.
Terlecky
,
J. F.
Dice
.
1995
.
Activation of a selective pathway of lysosomal proteolysis in rat liver by prolonged starvation.
Am. J. Physiol.
269
:
C1200
C1208
.
47
Kiffin
,
R.
,
C.
Christian
,
E.
Knecht
,
A. M.
Cuervo
.
2004
.
Activation of chaperone-mediated autophagy during oxidative stress.
Mol. Biol. Cell
15
:
4829
4840
.
48
Dice
,
J. F.
,
H. L.
Chiang
,
E. P.
Spencer
,
J. M.
Backer
.
1986
.
Regulation of catabolism of microinjected ribonuclease A. Identification of residues 7-11 as the essential pentapeptide.
J. Biol. Chem.
261
:
6853
6859
.
49
Bandyopadhyay
,
U.
,
S.
Kaushik
,
L.
Varticovski
,
A. M.
Cuervo
.
2008
.
The chaperone-mediated autophagy receptor organizes in dynamic protein complexes at the lysosomal membrane.
Mol. Cell. Biol.
28
:
5747
5763
.
50
Kaushik
,
S.
,
A. M.
Cuervo
.
2018
.
The coming of age of chaperone-mediated autophagy.
Nat. Rev. Mol. Cell Biol.
19
:
365
381
.
51
Fischer
,
J.
,
S.
Gutièrrez
,
R.
Ganesan
,
C.
Calabrese
,
R.
Ranjan
,
G.
Cildir
,
N. J.
Hos
,
J.
Rybniker
,
M.
Wolke
,
J. W. U.
Fries
, et al
.
2019
.
Leptin signaling impairs macrophage defenses against Salmonella Typhimurium.
Proc. Natl. Acad. Sci. USA
116
:
16551
16560
.
52
Komatsu
,
M.
,
H.
Kurokawa
,
S.
Waguri
,
K.
Taguchi
,
A.
Kobayashi
,
Y.
Ichimura
,
Y. S.
Sou
,
I.
Ueno
,
A.
Sakamoto
,
K. I.
Tong
, et al
.
2010
.
The selective autophagy substrate p62 activates the stress responsive transcription factor Nrf2 through inactivation of Keap1.
Nat. Cell Biol.
12
:
213
223
.
53
Ichimura
,
Y.
,
S.
Waguri
,
Y. S.
Sou
,
S.
Kageyama
,
J.
Hasegawa
,
R.
Ishimura
,
T.
Saito
,
Y.
Yang
,
T.
Kouno
,
T.
Fukutomi
, et al
.
2013
.
Phosphorylation of p62 activates the Keap1-Nrf2 pathway during selective autophagy.
Mol. Cell
51
:
618
631
.

The authors have no financial conflicts of interest.

This article is distributed under The American Association of Immunologists, Inc., Reuse Terms and Conditions for Author Choice articles.

Supplementary data