of February 16, 2018.This information is current as

InfectionListeriaRegulating Apoptosis and Crucial Roles of TNFAIP8 Protein in

ChenFayngerts, Derek S. Johnson, Zhaojun Wang and Youhai H.Yunwei Lou, Xiaohong Liang, Terry Cathopoulis, Svetlana Thomas P. Porturas, Honghong Sun, George Buchlis,

ol.1401987http://www.jimmunol.org/content/early/2015/05/06/jimmun

published online 6 May 2015J Immunol 

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2015 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on February 16, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Crucial Roles of TNFAIP8 Protein in Regulating Apoptosisand Listeria Infection

Thomas P. Porturas,* Honghong Sun,*,1 George Buchlis,*,1 Yunwei Lou,*,†

Xiaohong Liang,*,† Terry Cathopoulis,* Svetlana Fayngerts,* Derek S. Johnson,*

Zhaojun Wang,* and Youhai H. Chen*

TNF-a–induced protein 8 (TNFAIP8 or TIPE) is a newly described regulator of cancer and infection. However, its precise roles

and mechanisms of actions are not well understood. We report in this article that TNFAIP8 regulates Listeria monocytogenes

infection by controlling pathogen invasion and host cell apoptosis in a RAC1 GTPase-dependent manner. TNFAIP8-knockout mice

were resistant to lethal L. monocytogenes infection and had reduced bacterial load in the liver and spleen. TNFAIP8 knockdown in

murine liver HEPA1-6 cells increased apoptosis, reduced bacterial invasion into cells, and resulted in dysregulated RAC1 acti-

vation. TNFAIP8 could translocate to plasma membrane and preferentially associate with activated RAC1-GTP. The combined

effect of reduced bacterial invasion and increased sensitivity to TNF-a–induced clearance likely protected the TNFAIP8-knockout

mice from lethal listeriosis. Thus, by controlling bacterial invasion and the death of infected cells through RAC1, TNFAIP8

regulates the pathogenesis of L. monocytogenes infection. The Journal of Immunology, 2015, 194: 000–000.

Tumor necrosis factor-a–induced protein 8 (TNFAIP8 orTIPE), also known as SCC-S2, OXI-a, GG2-1, NDED,and MDC-3.13, is the first described member of the

TNFAIP8 family. It is an NF-kB–inducible protein upregulated inmetastatic head and neck squamous cell carcinoma cell lines andprotects cancer cells from TNF-a–induced apoptosis (1, 2). Over-expression in tumor cell lines also enhanced proliferation andmigration (3). TNFAIP8 was found to be a risk factor for non-Hodgkin’s lymphoma in humans and Staphylococcus aureus in-fection in mice (4). A few TNFAIP8-interacting partners havebeen identified, including activated Gai3 and karyopherin a2 (5).TNFAIP8L2, a closely related TNFAIP8 family protein, was re-ported to interact with RAC1 and to protect against Listeriamonocytogenes infection (6).L. monocytogenes is predominantly a food-borne pathogen,

with ∼10 cases/million inhabitants in most industrialized coun-tries. Despite having a lower incidence of infection than mostother food-borne pathogens, listeriosis carries a high risk formortality, ranging between 20 and 30%, and it accounts for 19%

of food-borne disease–related deaths in the United States (7, 8).L. monocytogenes is an intracellular Gram-positive bacterium thatinfects a number of cell types, including hepatocytes, neurons,and immune cells. Immune cell–mediated apoptosis of L. mono-cytogenes–infected cells, such as hepatocytes, is important forresolving infection (9, 10). One of the major mechanisms of im-mune defense against infectious intracellular microbes is toeliminate the infected cells by programmed cell death or apo-ptosis. We sought to determine the molecular pathways regulatedby TNFAIP8 and their relationship with L. monocytogenes in-fection and clearance. In this study, we show that TNFAIP8 sen-sitizes mice to lethal L. monocytogenes infection by potentiallyblocking apoptosis of infected cells and promoting the invasion byL. monocytogenes through RAC1. These results may provide newinsights into TNFAIP8’s regulation of cell death in listeriosis andcarcinogenesis.

Materials and MethodsAnimals

Wild-type (WT) C57BL/6 (B6) mice were purchased from The JacksonLaboratory. Tnfaip82/2 B6 mice were generated by germline gene tar-geting (11). All mice used in this study were housed under pathogen-freeconditions in the University of Pennsylvania Animal Care Facilities. Allanimal protocols used were preapproved by the Institutional Animal Careand Use Committee of the University of Pennsylvania.

Macrophage and neutrophil preparations

To generate bone marrow-derived macrophages (BMDMs), bone marrowcells were flushed from the femurs and tibias of donor mice. The RBCs werelysed with ACK solution (8.29 g NH4Cl, 1 g KHCO3, 37.2 mg Na2EDTA in1 l water). Cells were washed twice in ice-cold 13 Dulbecco’s PBS(DPBS) and cultured for 7 d in 30% L-929 cell culture supernatant and70% DMEM containing 10% (v/v) heat-inactivated FBS, 2 mM L-gluta-mine, and 100 U/ml penicillin/streptomycin (D10). Cells were washedtwice with cold DPBS and collected with 5 mM EDTA in DPBS. Aftercentrifugation, they were resuspended in D10 and rested for 24 h beforeexperimentation. BMDMs were .95% CD11b+ and F4/80+, as determinedby flow cytometry. Morphologically mature neutrophils were purified frommurine bone marrow by Percoll gradient centrifugation. Briefly, bonemarrow cells were harvested from mice using neutrophil isolation buffer(13 HBSS without Ca2+ and Mg2+ containing 0.25% BSA). After RBC

*Department of Pathology and Laboratory Medicine, University of PennsylvaniaPerelman School of Medicine, Philadelphia, PA 19104; and †Department of Immu-nology, Shandong University School of Medicine, Ji’nan 250012, People’s Republicof China

1H.S. and G.B. contributed equally to this work.

Received for publication August 5, 2014. Accepted for publication March 30, 2015.

This work was funded by National Institutes of Health Grants AI-077533, AI-050059, and GM-085112 (to Y.H.C.).

Address correspondence and reprint requests to Dr. Youhai H. Chen, Department ofPathology and Laboratory Medicine, University of Pennsylvania School of Medicine,713 Stellar-Chance Labs, 422 Curie Boulevard, Philadelphia, PA 19104-6160. E-mailaddress: yhc@mail.med.upenn.edu

Abbreviations used in this article: 7-AAD, 7-aminoactinomycin D; ALT, alanineaminotransferase; AST, aspartate aminotransferase; B6, C57BL/6; BHI, brain–heartinfusion; BMDM, bone marrow–derived macrophage; CHX, cycloheximide; DPBS,Dulbecco’s PBS; HA, hemagglutinin; KO, knockout; MOI, multiplicity of infection;PAK, p21-activated kinase; PBR, polybasic region; PGN, peptidoglycan; ROS, reac-tive oxygen species; shRNA, short hairpin RNA; TNFAIP8 (TIPE), TNF-a–inducedprotein 8; WT, wild-type.

Copyright� 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1401987

Published May 6, 2015, doi:10.4049/jimmunol.1401987 by guest on February 16, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

lysis, cells were layered on a 62% Percoll gradient. Following centrifu-gation at 1200 3 g for 30 min at room temperature, pelleted cells wereremoved and washed once with isolation buffer before being used in theexperiments. Neutrophil viability was .95%, as assessed by trypan bluestaining. Purity was typically 75–85%, as assessed by flow cytometrybased on forward and side scatter and CD11b and Gr1 staining.

Bone marrow chimeras

Bone marrow cells were flushed from the femurs and tibias of donor mice.The RBCs were lysed with ACK solution (8.29 g NH4Cl, 1 g KHCO3, 37.2mg Na2EDTA in 1 l water). Cells were washed twice and resuspended incold PBS. Recipient mice were sublethally irradiated twice with 500 rad,separated by 4 h. The irradiated mice received a total of 10 3 106 donorbone marrow cells by tail vein injection 1 or 2 h after irradiation. Micewere used 7–8 wk later for experiments.

Cell lines and plasmids

The HEK293T and Hepa1-6 cells were cultured in D10. Full-lengthTNFAIP8 cDNA was generated by PCR and cloned in-frame with anN-terminal Flag into vector pRK5. Human WT RAC1, RAC1 T17N, andRAC1 Q61L cDNAs were obtained from Addgene and subcloned intopRK5 with a Myc or hemagglutinin (HA) tag at the N terminus. Truncatedforms of RAC1 lacking the N-terminal aa 1–47 and C-terminal aa 162–192 or 189–192 were generated by PCR and cloned in-frame with anN-terminal HA tag into vector pRK5. cDNAs encoding TNFAIP8, WTRAC1, RAC1 T17N, and RAC1 Q61L were subcloned into the pEGFP-N3plasmid.

Plasmid DNA transfection and viral infection

293T cells were transfected with plasmid DNA using FuGENE 6 (Promega)reagent, according to the manufacturer’s instructions. For virus produc-tion, pLKO.1 (with puromycin resistance) with short hairpin RNA(shRNA)-Tnfaip8 (purchased from Open Biosystems) or shScr (a nonspecificscramble shRNA purchased from Addgene) fragments and packagingconstructs were cotransfected into 293T cells. After 24 and 48 h, virus-containing medium was filtered and used to infect the Hepa1-6 cell line inthe presence of 6.5 mg/ml Polybrene (Millipore). Infected cells were se-lected using puromycin (Sigma-Aldrich) to establish shRNA-Scr, shRNA-Tnfaip8-1 (SH1), and shRNA-Tnfaip8-2 (SH2) cell lines. Viral vectorswere produced from the pLKO.1 plasmid containing SH1 (59-AAAGGG-ATTGTACAAGGCAGC-39), SH2 (59-TTGAACTGATTGTTCCTGTGG-39),or shRNA-scr (59-CGAGGGCGACTTAACCTTAGG-39).

TNF-a stimulation and death assays

Hepa1-6 cell lines were serum starved for 16 h in DMEM containing 2 mML-glutamine and 100 U/ml penicillin/streptomycin. Cells were treated with10 ng/ml TNF-a for 0, 1, 5, 15, 30, or 60 min for protein lysates or with5 ng/ml TNF-a plus 20 ng/ml cycloheximide (CHX) for 7 or 16 h. Deathwas measured by flow cytometry using PE Annexin V and 7-amino-actinomycin D (7-AAD; BD Pharmingen) staining. For measuring deathfollowing RAC1 mutant transfections, EGFP+ cells were gated on forcomparison. To inhibit RAC1, cells were incubated with inhibitorZ62954982 (Millipore) at 100 mM for 30 min prior to experimentaltreatment.

Western blot

For Western blot analysis, cells were lysed for 20 min at 4˚C in 0.5%Nonidet P40 lysis buffer containing protease and phosphatase inhibitors.Cell debris was removed by centrifugation at 14,000 3 g for 20 min at4˚C. For subcellular fractionation, membrane and cytoplasmic proteinswere separated using a Qproteome Cell Compartment Kit (QIAGEN),according to the manufacturer’s instructions. The protein concentration ofthe lysates was determined by Bradford assay. Equal amounts of totalprotein were resolved by SDS-PAGE, transferred to membranes, andimmunoblotted with specific primary and secondary Abs, and the signalswere detected by chemiluminescence (Pierce). Primary Abs againstp-AKT(S473), AKT, Myc, and anti-Myc conjugated to HRP were pur-chased from Cell Signaling Technology and used according to the manu-facturer’s instructions. Anti-TNFAIP8 (1:500) was purchased fromProteintech Group, and anti-RAC1 (1:500) was purchased from Millipore.Anti-Flag (1:2000), anti–Flag-M2-HRP, anti–b-actin (1:3000), and anti-GAPDH (1:3000) Abs were purchased from Sigma-Aldrich. HRP-conjugated secondary anti-mouse and anti-rabbit IgG (1:1500) were pur-chased from GE Healthcare. The densitometric quantification of Westernblot signals was performed using ImageJ software.

Immunoprecipitation

Immunoprecipitation was performed using Dynabeads Protein G (Invi-trogen). In brief, 1.5 mg Dynabeads Protein G was coated with 5 mg specificAbs or Ig control for 1 h at room temperature with rotation. After removingunbound Ab, the bead–Ab complex was incubated with 500 ml cell lysatefor 4 h at 4˚C with rotation. The captured Dynabeads–Ab–Ag complex waswashed four times with PBS and boiled in 23 Laemmli buffer. The elutedproteins were fractionated by SDS-PAGE and detected by Western blot.

Loading of RAC1 with GDP and GTP

293T cells were transiently transfected with Myc-tagged RAC1 and Flag-tagged TNFAIP8 for 18 h. Cells were lysed in cell lysis buffer (50 mM Tris[pH 7.5], 10 mM EDTA, 0.2 M NaCl, 0.5% Nonidet P-40, and 13 proteaseinhibitors mixture; Roche). A total of 1 mM GDPbS or 0.2 mM GTPgS(Enzo Life Sciences) was loaded to transfected cell extracts. After 20 minof incubation at 30˚C, samples were immediately placed on ice, andMgCl2 was added to a final concentration of 10 mM to stop nucleotideexchange.

p21-Activated kinase pull-down assay

Hepa1-6 cells were serum starved for 16 h and then treated with 10 ng/mlTNF-a for 5 min. The cells were washed in PBS and lysed in PBD lysisbuffer (50 mM Tris [pH 7.5], 10 mM MgCl2, 0.2 M NaCl, 2% NonidetP-40, and 13 protease inhibitor mixture; Roche). The lysate was incubatedwith 20 mg p21-activated kinase (PAK)-GST protein beads (Cytoskeleton)for 30 min at 4˚C. After washing, protein on beads and in total cell lysateswas subjected to Western blot to determine the level of active Rac1.

F-actin assay

Hepa1-6 cells were plated on 12-well Nunc Multidishes with UpCellSurface (Thermo Scientific) and incubated overnight at 37˚C. Cells wereserum starved overnight and washed with PBS before resuspending in ice-cold PBS, according to the manufacturer’s protocol. Cells were fixed with4% paraformaldehyde for 10 min, permeabilized for 5 min in 0.1% TritonX-100, and stained with phalloidin-FITC (Sigma-Aldrich) for 30 min.Samples were analyzed by flow cytometry, and mean fluorescence inten-sity was calculated using FlowJo software.

Microbial strains and infection

WT L. monocytogenes (10403s) were provided by H. Shen and Y. Paterson(University of Pennsylvania). L. monocytogenes and S. aureus (ATCC29213) were grown at 37˚C in brain–heart infusion (BHI) medium (BectonDickinson) and Columbia medium with 2% NaCl, respectively. For allassays, midlog-phase bacteria were used. Bacterial in vitro infection wasassayed in six-well plates. A total of 7.5 3 105 Hepa1-6 cells was seededin each well, followed by culturing overnight in DMEM with 10% FBS.Cells were serum starved for 18–24 h and infected with L. monocytogenesat a multiplicity of infection (MOI) of 50 in antibiotic-free media. Sixtyminutes after the inoculation, cells were washed three times with PBS,and fresh medium containing gentamicin (150 mg/ml) was added. At1.5 h postinfection, cells were lysed with 0.1% Triton in PBS, and serialdilutions of the homogenate were plated on BHI agar plates (BectonDickinson). The colonies were counted 24 h later. For in vivo bacterialinfection, L. monocytogenes was grown in BHI medium until the absor-bance at 600 nm reached 0.1 of OD. S. aureus was grown in Columbiamedia with 2% NaCl. Six to seven-week-old WT and TNFAIP8-knockout(KO) mice were infected i.v. with 2 3 105 L. monocytogenes in 200 mlPBS or 2 3 107 S. aureus in 200 ml saline. For measurement of thebacterial burden in liver and spleen, mice were sacrificed 72 h after in-oculation, organs were homogenized in 0.1% Triton in PBS, and serialdilutions of the homogenate were plated on BHI agar plates. The colonieswere counted 24 h later. Serum alanine aminotransferase (ALT) and as-partate aminotransferase (AST) levels were determined using the InfinityALT or AST liquid stable reagent (Thermo).

Phagocytosis assay

To prepare apoptotic cells for phagocytosis assay, thymocytes were har-vested from 3–4-wk-old B6 mice, loaded with 10 mM CFSE (MolecularProbes) for 10 min at 37˚C, and washed twice in PBS. Apoptosis wasinduced by incubation at 37˚C in 5% CO2 for 5 h in the presence of 5 mMdexamethasone (Sigma-Aldrich). After dexamethasone treatment, cellswere washed three times with PBS and resuspended in DMEM with 2%FBS. This treatment routinely yielded .70% apoptotic thymocytes, asmeasured by annexin V staining. Phagocytosis assay was performed ina 12-well nontissue culture–treated plate (BD Falcon). A total of 1 3 106

2 TNFAIP8 IN APOPTOSIS AND LISTERIOSIS

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

BMDMs was seeded in each well, followed by culturing overnight inDMEM with 2% FBS. CFSE+ apoptotic thymocytes were added at a ratioof 5:1, and centrifugation was performed at 500 3 g for 2 min to syn-chronize binding and internalization. After 30 min of incubation at 37˚C,plates were washed rapidly two times with ice-cold PBS, and cells werecollected with 5 mM EDTA-PBS. Cells were then fixed with 2% para-formaldehyde in PBS, stained with allophycocyanin-conjugated anti-F4/80, and analyzed by flow cytometry. Gates were set for macrophages inforward scatter/side scatter dot blots. Experiments using fluorescent latexbeads (2 mm; Sigma-Aldrich) were performed in a similar fashion. Forbacterial phagocytosis assay, L. monocytogenes were washed twice withsterile PBS and incubated (at 1 3 109/ml bacteria) in 2 mM CFSE for 20–30 min under constant shaking at 37˚C. CFSE-labeled bacteria werewashed three times with PBS before being used. Live bacteria were fed toBMDMs at a ratio of 10:1 in DMEM with 10% serum.

Chemotaxis assay

Primary neutrophils were loaded with 5 mM calcein AM in PBS supple-mented with 2% FBS for 30 min at 37˚C. After three washes with PBS,neutrophils were resuspended to 2 3 106 cells/ml PBS. ChemoTx plates(Neuro Probe) were loaded with 50 ml cell over each 5.7-mm-diameterchamber with 5-mm pores. The lower chambers were loaded with 30 mlPBS containing the specified concentration of IL-8. Neutrophils were in-cubated for 1 h, the top chambers were washed off, and the plates wereread in a fluorescent plate reader. Chemotactic index was calculated as thenumber of neutrophils in the lower chambers of chemokine-containingwells over that in the wells with no chemokines.

ELISA

Sera were collected from WT and TNFAIP8-KO mice 72 h after L. mono-cytogenes infection and kept at 280˚C. Abs used in ELISA were pur-chased from BD Pharmingen and eBioscience, including purified andbiotinylated rat anti-mouse TNF-a. Quantitative ELISA was performedusing paired mAbs specific for corresponding cytokines, according to themanufacturer’s recommendations.

Statistical analysis

Quantitative data are presented as means 6 SEM of two or three experi-ments. Survival curves were plotted via the Kaplan–Meier method andcompared by the log-rank test. The two-tailed Student t test was used forall other results, and p , 0.05 was considered statistically significant. Allstatistical analyses were performed using GraphPad Prism software.

ResultsTNFAIP8-KO mice are resistant to lethal L. monocytogenesinfection

It was reported that TIPE2-deficient mice are resistant toL. monocytogenes infection (6). We hypothesized that TNFAIP8-KO mice might have a similar phenotype. To test this, TNFAIP8-KO and WT B6 mice were injected i.v. with a lethal dose ofL. monocytogenes. As expected, TNFAIP8-KO mice were resis-tant to death (Fig. 1A) and experienced reduced weight loss duringinfection (Fig. 1B). Next, we examined the bacterial load in theliver and spleen 3 d postinfection and found reduced bacterial loadin the KO mice (Fig. 1C). We assessed liver damage by measuringserum levels of ALT and AST, but we found no significant dif-ferences between WT and KO mice (Fig. 1D). Serum TNF-a andIL-6 levels also were assayed. TNF-a and IL-6 are important forcontrolling L. monocytogenes, and mice deficient in either of themhave increased bacterial burden (12). Although the KO mice hadless TNF-a and more IL-6, the differences were not statisticallysignificant (Fig. 1E). Therefore, TNFAIP8-KO mice are resistantto lethal L. monocytogenes infection, independent of serumcytokines and hepatocellular toxicity.

Immune system function does not sufficiently explain theresistance to lethal L. monocytogenes in TNFAIP8-KO mice

To test whether TNFAIP8-KO immune cells are responsible forthe resistance to L. monocytogenes infection, we produced bonemarrow chimeric mice using both WT and KO irradiated mice as

recipients of WT bone marrow. After their immune cells werereplenished, the recipient mice were challenged with a lethal doseof L. monocytogenes. We once again found that TNFAIP8-KOmice exhibited significant resistance to death (Fig. 2A). Next,we tested whether there was a difference in mouse survival withanother Gram-positive pathogen. We chose S. aureus because theTnfaip8 gene was reported to be a risk factor for its infection andit is an extracellular pathogen (4). TNFAIP8-KO and WT micewere infected i.v. with a lethal dose of S. aureus, but there was nodifference in mouse survival (Fig. 2B). To investigate whetherthere were any differences in innate immune cell function, weisolated primary neutrophils from bone marrow and assessedneutrophil chemotaxis. We found no significant difference be-tween WT and KO cells with regard to chemotaxis toward IL-8 ina Boyden chamber assay (Fig. 2C). Although IL-8 is not expressedin mice, murine neutrophils express the functional IL-8Rb(CXCR2), and IL-8 was used previously to induce chemotaxisin mice (13, 14). We next studied the phagocytic capacity ofBMDMs and found that there were no differences between WTand KO cells in the phagocytosis of killed CFSE-labeled thymo-cytes, fluorescently labeled beads, or L. monocytogenes (Fig. 2D).To determine whether TNFAIP8 deficiency affects the productionof TNF-a and IL-6 in response to peptidoglycan (PGN), a majorcomponent of Gram-positive bacteria, we injected PGN into WTand TNFAIP8-KO mice and checked the serum levels of thecytokines. Again, we did not find any significant differences be-tween the two groups (Fig. 2E). Similarly, BMDMs did not showa difference in the production of TNF-a and IL-6 upon stimulation

FIGURE 1. TNFAIP8-KO mice are resistant to lethal L. monocytogenes

infection. WT (n = 5) and TNFAIP8-KO mice (n = 7) were infected with

23 105 CFU of L. monocytogenes through the tail vein. (A) Kaplan–Meier

survival curve after L. monocytogenes injection (p = 0.01). (B) Changes in

body weight following infection (p = 0.04). Mice were sacrificed at day 3

postinfection, and their livers, spleens, and sera were collected. (C) Spleen

and liver bacterial load was measured by a colony-forming assay. (D)

Serum levels of AST and ALT. (E) Serum TNF-a and IL-6 levels were

measured by ELISA. The results are representative of at least two inde-

pendent experiments. *p , 0.05.

The Journal of Immunology 3

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

with PGN in vitro (Fig. 2F). Taken together, these results suggestthat the immune system does not play a primary role in the re-sistance to lethal L. monocytogenes infection in TNFAIP8-KOmice.

TNFAIP8 regulates L. monocytogenes infection and the deathof infected cells in vitro

Because differences in immune function did not adequately explainthe resistance to lethal L. monocytogenes infection in TNFAIP8-KO mice, we next investigated the roles of TNFAIP8 in nonim-mune cells involved in bacterial infection. Because hepatocytesare a major target of L. monocytogenes infection, we generatedTNFAIP8-knockdown Hepa1-6 murine liver cells. Using a lenti-viral vector expressing either a scrambled shRNA sequence orsequences targeting TNFAIP8, we created several knockdown celllines and determined the protein levels of TNFAIP8 and AKT(Fig. 3A). We found that shRNA–tnfaip8-1 (SH1) was a nearlycomplete knockdown, with increased AKT phosphorylation at

serine 473. AKT phosphorylation at serine 473 is regulated byMTORC2 and PI3K and is important in regulating apoptotic celldeath, and PI3K itself is important in L. monocytogenes cell in-vasion (15, 16).Next, we tested whether L. monocytogenes invasion of Hepa1-6

cells was affected by TNFAIP8 depletion. After allowing thebacteria to invade cells for 1 h, we found that the knockdown cellshad reduced bacterial burden, with or without TNF-a treatment,and TNF-a did not provide any protective effect against bacterialinvasion (Fig. 3B). TNF-a is important in protecting mice againstmany bacterial pathogens. Hepatocytes produce TNF-a upon ex-posure to L. monocytogenes, and its depletion sensitizes mice todeath by L. monocytogenes (12, 17). Next, we tested whetherTNFAIP8 knockdown had an effect on L. monocytogenes–inducedcell death. We incubated the cells with L. monocytogenes over-night, with or without TNF-a, and found that the TNFAIP8-knockdown cells showed enhanced sensitivity toward cell deathonly when treated with both L. monocytogenes and TNF-a(Fig. 3C). Finally, we assayed whether there was a difference inTNF-a–induced cell death by treating cells with TNF-a and CHXand found increased sensitivity in the knockdown cell lines(Fig. 3D). Our results suggest that the resistance to lethalL. monocytogenes infection in mice could be caused by the re-sistance of host cells to bacterial invasion and increased sensitivity

FIGURE 2. Immune system function does not sufficiently explain the

resistance to lethal L. monocytogenes of TNFAIP8-KO mice. (A) Bone

marrow chimeras were produced by irradiating WT (n = 6) and TNFAIP8-

KO (n = 6) mice and transferring WT bone marrow into them. Mice were

given 8 wk to reconstitute their immune system before being infected with

2 3 105 CFU of L. monocytogenes. The difference in survival between the

two groups was statistically significant (p = 0.02). (B) WT (n = 5) and

TNFAIP8-KO (n = 7) mice were challenged with 2 3 107 CFU of

S. aureus. The difference in survival between the two groups was not

statistically significant. (C) Calcein AM–labeled primary neutrophils were

incubated for 1 h in a Boyden chamber filled or not with 50 or 500 ng/ml of

IL-8. The rate of chemotaxis was determined using a fluorescent plate

reader. (D) BMDMs were incubated with 2-mm fluorescent beads at a ratio

of 1:20, apoptotic CFSE-labeled thymocytes at a ratio of 1:5, or CFSE-

labeled L. monocytogenes at a ratio of 1:10 for 30 min. The rate of

phagocytosis was measured by flow cytometry. (E) WT and TNFAIP8-KO

mice were injected i.v. with PGN (10 mg/kg), and serum was collected

after 8 h to assay TNF-a and IL-6 levels by ELISA. (F) BMDMs were

treated with PGN (10 mg/ml) for 8 h, and TNF-a and IL-6 levels were

assayed in the supernatant. The results are representative of at least two

independent experiments.

FIGURE 3. TNFAIP8 regulates L. monocytogenes infection and the

death of infected cells in vitro. Hepa1-6–knockdown cells were generated

using the pLKO.1 lentiviral vector to express either a scrambled shRNA

(Scr) or one of the two antisense sequences targeting TNFAIP8 (SH1 and

SH2). (A) Western blot was performed for the indicated proteins in the

lysates of untransfected control (UT) and shRNA-treated Hepa1-6 cells.

(B) L. monocytogenes infection of control and TNFAIP8-knockdown

Hepa1-6 cells was measured in a colony formation assay. Cells were

infected with 50 MOI of L. monocytogenes for 1 h, with or without

50 ng/ml of TNF-a. Cells were washed and treated with 150 mg/ml

gentamicin for 30 min to kill extracellular bacteria before being lysed to

release the intracellular bacteria. (C) Hepa1-6 TNFAIP8-knockdown and

control cells were challenged with L. monocytogenes (L) at 60 MOI, with

or without 50 ng/ml of TNF-a (T) for 16 h. Death was quantified by flow

cytometry after annexin V staining. (D) Hepa1-6–knockdown and con-

trol cell lines were treated or not with 5 ng/ml of TNF-a and 20 ng/ml of

CHX for 6 h. Cell death was determined by flow cytometry after annexin

V and 7-AAD staining. The results are representative of at least two

independent experiments. **p , 0.01, *** p , 0.001. NT, not treated

with TNF-a.

4 TNFAIP8 IN APOPTOSIS AND LISTERIOSIS

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

to cell death following infection, thereby reducing bacterial loadin the organism.

TNFAIP8 regulates RAC1 activation by TNF-a

RAC1 is an important regulator of cell death following TNF-astimulation, and TNFAIP8 was found to protect against TNF-a–induced cell death (2, 3, 5, 18, 19). RAC1 also was found toregulate L. monocytogenes invasion into endothelial cells (6, 20).To test whether there is a difference in RAC1 activation, PAK-GST beads were used to pull down RAC1. PAK only binds toactive RAC1-GTP and is routinely used to assay for RAC1 ac-tivity. Knockdown and control cells were stimulated with TNF-afor 5 min, and the amount of RAC1 pulled down with PAK-GSTbeads was compared (Fig. 4A). We found dysregulated RAC1activation in TNFAIP8-knockdown cells, with higher basal levelsof RAC1-GTP, but limited induction following TNF-a stimulation.Next, we checked whether there were any differences in RAC1trafficking to the membrane and found impaired RAC1 transloca-tion following TNF-a stimulation (Fig. 4B). In control cells,TNFAIP8 seemed to localize primarily to the cytosol, but, surpris-ingly, there was more TNFAIP8 on the membrane than in the cy-tosol in the knockdown cells. To determine the consequence ofincreased basal RAC1 levels, we assayed F-actin by flow cytometryand found increased levels in TNFAIP8-knockdown cells (Fig. 4C).To test whether the differences in knockdown cell death are

RAC1 dependent, we transfected cells with the constitutively activeRAC1-61L or inactive RAC1-17N and treated them with TNF-aand CHX (Fig. 4D). We found that the dominant-negative mutantpartially rescued cell death, whereas the constitutively activemutant greatly sensitized the control cells to death but had a lim-

ited effect on the knockdown cells. We next tested the relevance ofRAC1 signaling to L. monocytogenes invasion of the knockdowncells. We infected TNFAIP8-knockdown and control cells withL. monocytogenes, with or without RAC1 inhibitor treatment, andfound that the inhibitor effectively protected cells from bacterialinvasion and negated any differences between control andTNFAIP8-knockdown cells (Fig. 4E). Taken together, these resultsindicate that RAC1 activation is dysregulated in TNFAIP8-knockdown cells, which likely contributes to the altered re-sponse to TNF-a–induced cell death and L. monocytogenes in-vasion.

TNFAIP8 associates with RAC1 with a higher affinity towardactive RAC1-GTP

We first tested whether Flag-tagged TNFAIP8 associates withendogenous RAC1. We transfected 293T cells with a plasmidencoding Flag-tagged TNFAIP8 and immunoprecipitated celllysates with anti-Flag Ab or control IgG and blotted for RAC1and Flag-TNFAIP8 (Fig. 5A). We found that endogenous RAC1coimmunoprecipitated with TNFAIP8. We then tested whetherthis interaction was dependent on a particular region of RAC1. Wecreated several HA-tagged RAC1 mutants and transfected theminto 293T cells along with Flag-tagged TNFAIP8. We pulled downFlag and blotted for HA and found that RAC1 association wasimpaired with the D162–192 RAC1 mutant, which lacks the poly-basic region (PBR) and CAAX motif (Fig. 5B). The PBR iscritical for RAC1 interaction with certain proteins and propercellular localization (21, 22). The CAAX motif is responsible forposttranslational modification and is crucial for the proper func-tion of RAC1 (23, 24).

FIGURE 4. TNFAIP8 and RAC1 in L. monocytogenes infection and cell death. (A) The levels of RAC1-GTP were determined by a PAK-GST pull-down

assay for control and TNFAIP8-knockdown Hepa1-6 cells stimulated or not with 50 ng/ml of TNF-a for 5 min. Relative band intensities: 1, 1.5, 1.7, and

1.54 for Scr and SH1 with no treatment, and Scr and SH1 with TNF-a treatment, respectively. (B) Control and TNFAIP8-knockdown Hepa1-6 cells were

stimulated with 50 ng/ml of TNF-a for 0, 2, 5, or 15 min. Lysates were fractionated into membrane (M) and cytoplasmic (C) portions, and TNFAIP8,

RAC1, and actin were detected by Western blot. Relative band intensities: 1, 1.61, 2.86, 1.84, 1.72, 1.42, 1.91, and 2.04 for Scr with 0, 2, 5, and 15 min of

TNF-a treatment and SH1 with 0, 2, 5, and 15 min of TNF-a treatment, respectively. Band intensities were determined by ImageJ software. (C) Control and

TNFAIP8-knockdown Hepa1-6 cells were stained with phalloidin-FITC to quantify their total levels of F-actin by flow cytometry. (D) Control and

TNFAIP8-knockdown Hepa1-6 cells were transiently transfected with p-EGFP vectors containing either the EGFP cDNA alone (E) or EGFP plus RAC1-

17N (R-) or RAC1-61L (R+) cDNAs. Cells were treated with 5 ng/ml of TNF-a and 20 ng/ml of CHX for 6 h. EGFP+ cells were gated by flow cytometry,

and cell death was detected by annexin V and 7-AAD staining. (E) Control and TNFAIP8-knockdown Hepa1-6 cells were infected with 50 MOI of

L. monocytogenes for 1 h, with or without RAC1 inhibitor II Z62954982 at 100 mM. Cells were washed and treated with 150 mg/ml gentamicin for 30 min

to kill extracellular bacteria before being lysed to release the intracellular bacteria. The results are representative of at least two independent experiments.

*p , 0.05, **p , 0.01, ***p , 0.001.

The Journal of Immunology 5

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

To test whether TNFAIP8–RAC1 interaction is dependent onthe activation status of RAC1, we loaded 293T lysate-express-ing Myc-tagged RAC1 and Flag-tagged TNFAIP8 with non-hydrolyzable forms of GTP (GTPgS) and GDP (GDPbS). Wefound increased TNFAIP8 association with GTPgS-loaded RAC1,suggesting that TNFAIP8 preferentially associates with the activeform of RAC1 (Fig. 5C). These results demonstrate that TNFAIP8associates with active RAC1 and that this association may beresponsible for the TNFAIP8 effect in L. monocytogenes infection.

DiscussionThe role of TNFAIP8 in cell death appears to be condition de-pendent. In most tumor cells, a protective effect of TNFAIP8 isreported; however, in thymocytes, TNFAIP8 was found to promoteglucocorticoid-induced apoptosis (25). The molecular pathwaysof cell death that are regulated by TNFAIP8 are still largely un-known. Previous work suggests that TNFAIP8 protects againstdeath through its “death effector domain,” but that hypothesis hasbeen largely debunked by the structural analysis of TIPE2, whichrevealed a novel fold instead of a death effector domain (3, 26).Other investigators suggested that TNFAIP8 works through Ga(i)proteins and showed that TNFAIP8 is important for the function ofdopamine-D2 short receptors in protecting against TNF-a–in-duced cell death (5). Recently, TNFAIP8L1, another TNFAIP8family member, was implicated in activating mTOR to protectagainst oxidative stress in dopamine neurons by blocking auto-phagy (27). The mTOR protein itself plays an important role inpromoting cancer, protecting against cell death, and increasingsusceptibility to lethal L. monocytogenes infection, and it also wasfound to be regulated by the small GTPase RAC1 (28–30). Weshow in this study that TNFAIP8 associates with RAC1 and reg-ulates TNF-a–induced cell death and lethal L. monocytogenesinfection in mice. We propose that it does so by regulating RAC1signaling (Fig. 6).L. monocytogenes pathogenesis is dependent on a number of

factors related to both immune and nonimmune systems. Hepa-tocytes are quickly invaded by L. monocytogenes upon infection,and apoptotic cell death is important in releasing TNF-a and othercytokines to trigger an immune response to the infected liver (9,17, 31, 32). Neutrophils are attracted to the infected organs tophagocytose and kill bacteria. If the pathogen is not controlled,L. monocytogenes will escape to other organs, and the organism

can die of sepsis. Protecting immune cells from death is an im-portant aspect in preventing lethal L. monocytogenes infection (33,34); however, promoting cell death in hepatocytes or blockingcellular invasion might be equally important for the survival of theorganism (17, 20, 35).Our results show that TNFAIP8-KO animals are resistant to

lethal L. monocytogenes infection with decreased bacterial load inthe liver and spleen, similar to what was observed with TIPE2 (6).However, unlike TIPE2-KO mice that show heightened immunityto L. monocytogenes, there was little difference in phagocytosis,chemotaxis, or cytokine production between WT and TNFAIP8-KO innate immune cells. TNF-a and IL-6 are important in theresistance to L. monocytogenes infection, and both were studiedpreviously in TIPE2-KO mice (2, 6, 12). In addition to thesecytokines, IFN-g is very important for immunity to L. mono-cytogenes (31, 36). However, due to limitations in serum quanti-ties collected, we could not investigate IFN-g in this study.Despite reports suggesting a role for TNFAIP8 in S. aureus in-fection, we found no difference in survival between WT andTNFAIP8-KO mice (4). This suggests that TNFAIP8-KO mice arenot resistant to extracellular pathogens and that the immune sys-tem is no more capable of handling pathogens than is the one inWT mice. Rather, the nonimmune cells may be more important inprotecting the host against intracellular pathogens.Apoptosis of host cells is an important defense mechanism

against intracellular pathogens like L. monocytogenes (10). He-patocyte apoptosis may inhibit L. monocytogenes spread withinthe liver and allow access of the pathogen to professional phag-ocytes (32). Previous work focused on leukocyte-mediated celldeath of hepatocytes, and it was discovered that depletion of Fasor the downstream caspase-8 enzyme in hepatocytes reducedCD8+ T cell–mediated cell death and resulted in increased bac-terial burden in the liver (37, 38). The leukocytes responsible forlysis of hepatocytes during the early stages of infection are notclear, with both neutrophils and macrophages likely involved (39,40). Therefore, we decided to investigate L. monocytogenes in-fection in the nonimmune hepatoma cell line Hepa1-6. We foundthat TNFAIP8-knockdown Hepa1-6 cells were more sensitive todeath by L. monocytogenes when treated with TNF-a and wereresistant to L. monocytogenes invasion. TNF-a is critical forprotecting against lethal L. monocytogenes infection, which canblock certain aspects of TNF-a signaling to protect itself (12, 41).

FIGURE 5. TNFAIP8 associates with RAC1 with a higher

affinity toward active RAC1-GTP. (A) 293T cells were trans-

fected with the pRK5 expression plasmids containing Flag-

tagged TNFAIP8. Anti-FLAG Ab or control mouse IgG was

used to immunoprecipitate the Flag-tagged TNFAIP8 along

with endogenous RAC1, and Western blotting was performed

using the indicated Abs. (B) HA-tagged RAC1 mutants were

coexpressed with Flag-tagged TNFAIP8 in 293T cells, and

immunoprecipitation (IP) was performed using anti-Flag.

Western blots were developed using anti-HA Ab to identify

RAC1 mutants coimmunoprecipitated with TNFAIP8. (C)

293T cells were cotransfected with Flag-tagged TNFAIP8 and

Myc-tagged RAC1 or empty pRK5 vector. Total cell lysates

were treated or not with GTPgS or GDPbS, and anti-Myc Ab

was used to immunoprecipitate RAC1. TNFAIP8 and RAC1

were determined by Western blotting. The results are repre-

sentative of at least two independent experiments.

6 TNFAIP8 IN APOPTOSIS AND LISTERIOSIS

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Therefore, it stands to reason that TNFAIP8-KO mice are pro-tected against lethal L. monocytogenes infection by first reducinghepatocyte invasion and second by increasing apoptosis of infec-ted hepatocytes in response to TNF-a, thus blocking a productiveinfection of the pathogen.We observed an increase in p-AKT at serine 473 in TNFAIP8-

knockdown cells, suggesting that mTOR also may be regulated byTNFAIP8. RAC1 has been implicated in regulating mTOR com-plexes by its interaction with mTORC1 and mTORC2 proteins(30). In addition, TNFAIP8L2 was found to associate with RAC1(6). This led us to investigate whether TNFAIP8 effector functionsare dependent on RAC1 signaling, because both RAC1 and mTORare implicated in L. monocytogenes invasion of cells and apoptosis(16, 20, 29).After testing several conditions using TNFAIP8-knockdown

cells, we concluded that TNFAIP8’s regulation of cell death isRAC1 dependent. By introducing the dominant-negative mutant ofRAC1, TNF-a–induced cell death in TNFAIP8-knockdown cellsis partially rescued. Furthermore, the constitutively active mu-tant had a stronger effect on the control cells than it did on theknockdown cells, presumably because the RAC1 signaling alreadywas heightened in knockdown cells. We observed increasedRAC1-GTP at basal levels in knockdown cells and enrichment ofTNFAIP8 in the membrane fraction compared with the controls.The localization of TNFAIP8 in knockdown cells suggests that themembrane association is crucial for proper cell function and that itmay be related to the increased RAC1 activation. In addition, wefound increased levels of F-actin in knockdown cells at the basalstate, supporting the heightened level of RAC1-GTP in these cells.This led us to the conclusion that TNFAIP8-knockdown cells aremore susceptible to TNF-a–induced cell death because there isinsufficient TNFAIP8 present to inhibit RAC1.Bacterial entry into knockdown cells is inhibited, despite in-

creased RAC1 activation and reports that RAC1 activation isnecessary for bacterial entry (20, 42). The RAC1 inhibitorinhibited bacterial invasion of both knockdown and control cellsand eliminated the differences in bacterial counts, suggesting thatRAC1 is responsible for the differences in bacterial invasion. Apossible explanation for reduced bacterial invasion of knockdowncells in the face of elevated RAC1 activation is that RAC1 isactivated nonspecifically throughout the cellular membrane inTNFAIP8-deficient cells, increasing reactive oxygen species

(ROS) production and activating molecular pathways that do notrequire localized activation. Functions that require localized ac-tivation, like the internalization of an L. monocytogenes cell ona host membrane, may be inhibited because there is too muchbackground RAC1 activation to efficiently recruit the requiredproteins to a specific location and internalize the bacterium. Analternative explanation is that infected TNFAIP8-knockdown cellsare dying rapidly, reducing the total number of bacteria protectedfrom the gentamicin treatment. However, this explanation seemsunlikely because of the short incubation time with L. mono-cytogenes for the infection experiments, and no difference in deathinduced by L. monocytogenes was observed after 24 h comparedwith control cells.There are several forms of endocytosis, and L. monocytogenes

was shown to take advantage of clathrin-dependent MET recep-tor–mediated endocytosis to gain entry into the cell (43). Contraryto the differences observed with L. monocytogenes invasion ofthe Hepa1-6 cell lines, phagocytosis by TNFAIP8-KO BMDMsshowed no abnormalities. One possible explanation is that the lackof TNFAIP8 in the KO macrophages is compensated for by thepresence of TIPE2. TIPE2 protein is undetectable in Hepa 1-6cells (data not shown) but is expressed in macrophages where itregulates phagocytosis (6). A second potential explanation is thatthe molecular mechanism for phagocytosis of L. monocytogenesin macrophages is sufficiently different from the endocytosis ofL. monocytogenes in Hepa 1-6 cells that TNFAIP8 deficiency canaffect one but not the other.Finally, we showed that TNFAIP8 associates with RAC1

by coimmunoprecipitation. We were forced to use transientlyexpressed TNFAIP8 because of the very weak binding affinity ofthe commercial TNFAIP8 Abs. We further found that the TNFAIP8association is dependent on the C terminus of RAC1, whichcontains the PBR and CAAX motif. These two regions are re-sponsible for protein interactions, membrane localization signal-ing, and posttranslational modifications. Our finding that theassociation is dependent on whether RAC1 is primed with GTP orGDP suggests that RAC1 interaction with TNFAIP8 is enhancedwith RAC1 activation. Understanding the molecular interaction ofTNFAIP8 with RAC1 is important in identifying therapeutic targetsfor treating a wide variety of diseases, including bacterial infectionsand cancer.We propose that TNFAIP8 regulates TNF-a–induced cell death

by its inhibition of RAC1 (Fig. 6). The TNF-a signaling cascadeactivates prodeath pathways through caspase cleavage and ROSproduction by the NADPH oxidase complex, as well as pro-survival pathways through the activation of the NF-kB transcrip-tion factor. Inhibition of RAC1 and the associated NADPHoxidase activation were documented to protect cells against bothapoptotic and necrotic cell death following TNF-a stimulation(18, 19). TNFAIP8 expression is reported to be upregulated byTNF-a signaling through NF-kB (2, 44). Our findings suggest thatNF-kB can protect against cell death by preventing ROS pro-duction through RAC1 inhibition by the expression of TNFAIP8.We also propose that RAC1 regulation by TNFAIP8 is important

for L. monocytogenes invasion of host cells. L. monocytogenesinternalin A and B surface proteins bind to MET (also known asthe hepatocyte growth factor receptor), b-integrin, and E-cadherinreceptors to activate RAC1 and induce F-actin cytoskeletalremodeling that allow for bacterial entry into the cells (42). Wehypothesize that TNFAIP8 functions to reduce background RAC1activation, allowing for precise control of RAC1 signaling. In-sufficient TNFAIP8 likely interferes with this aspect of bacterialinfection by preventing localized RAC1 signaling at the site ofbacterial entry, thus preventing the recruitment of F-actin and

FIGURE 6. Model of TNFAIP8 molecular function in cell death. TNF-a

activates TNFR, which results in RAC1 and NF-kB activation. RAC1 then

activates the NADPH oxidase complex to generate ROS that promote cell

death. The NF-kB transcription factor promotes TNFAIP8 expression to

inhibit RAC1 activation and protect against cell death.

The Journal of Immunology 7

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

inhibiting bacterial internalization. In conclusion, our results in-dicate that TNFAIP8 regulates RAC1 signaling during TNF-aRactivation and L. monocytogenes invasion of cells and that thisregulation is crucial for controlling cell death and lethal L. mono-cytogenes infection of mice.

AcknowledgmentsWe thank Hao Shen, Yvonne Paterson, Ji Qi, Qingguo Ruan, and George

Luo for reagents and/or valuable advice. We also acknowledge Drs.

Margaret M. Chou, Nicola J. Mason, Robert H. Vonderheide, and Kenneth

S. Zaret for helpful comments and discussions.

DisclosuresThe authors have no financial conflicts of interest.

References1. Patel, S., F. H. Wang, T. L. Whiteside, and U. Kasid. 1997. Identification of

seven differentially displayed transcripts in human primary and matched meta-static head and neck squamous cell carcinoma cell lines: implications in me-tastasis and/or radiation response. Oral Oncol. 33: 197–203.

2. Kumar, D., T. L. Whiteside, and U. Kasid. 2000. Identification of a novel tumornecrosis factor-alpha-inducible gene, SCC-S2, containing the consensus se-quence of a death effector domain of fas-associated death domain-like inter-leukin- 1beta-converting enzyme-inhibitory protein. J. Biol. Chem. 275: 2973–2978.

3. Kumar, D., P. Gokhale, C. Broustas, D. Chakravarty, I. Ahmad, and U. Kasid.2004. Expression of SCC-S2, an antiapoptotic molecule, correlates with en-hanced proliferation and tumorigenicity of MDA-MB 435 cells. Oncogene 23:612–616.

4. Ahn, S.-H., H. Deshmukh, N. Johnson, L. G. Cowell, T. H. Rude, W. K. Scott,C. L. Nelson, A. K. Zaas, D. A. Marchuk, S. Keum, et al. 2010. Two genes onA/J chromosome 18 are associated with susceptibility to Staphylococcusaureus infection by combined microarray and QTL analyses. PLoS Pathog.6: e1001088. Available at: http://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1001088.

5. Laliberte, B., A. M. Wilson, H. Nafisi, H. Mao, Y. Y. Zhou, M. Daigle, andP. R. Albert. 2010. TNFAIP8: a new effector for Galpha(i) coupling to reducecell death and induce cell transformation. J. Cell. Physiol. 225: 865–874.

6. Wang, Z., S. Fayngerts, P. Wang, H. Sun, D. S. Johnson, Q. Ruan, W. Guo, andY. H. Chen. 2012. TIPE2 protein serves as a negative regulator of phagocytosisand oxidative burst during infection. Proc. Natl. Acad. Sci. USA 109: 15413–15418.

7. Mateus, T., J. Silva, R. L. Maia, and P. Teixeira. 2013. Listeriosis duringPregnancy: A Public Health Concern. ISRN Obstet. Gynecol. 2013: 851712.

8. Hernandez-Milian, A., and A. Payeras-Cifre. 2014. What is new in listeriosis?Biomed. Res. Int. 2014: 358051.

9. Vazquez-Boland, J. A., M. Kuhn, P. Berche, T. Chakraborty, G. Domınguez-Bernal, W. Goebel, B. Gonzalez-Zorn, J. Wehland, and J. Kreft. 2001. Listeriapathogenesis and molecular virulence determinants. Clin. Microbiol. Rev. 14:584–640.

10. Kaufmann, S. H. 1993. Immunity to intracellular bacteria. Annu. Rev. Immunol.11: 129–163.

11. Sun, H., Y. Lou, T. Porturas, S. Morrissey, G. Luo, J. Qi, Q. Ruan, S. Shi, andY. H. Chen. 2015. Exacerbated experimental colitis in TNFAIP8-deficient mice.J. Immunol. 194: XXXX–XXXX.

12. Miura, T., S. Nishikawa, S. Sasaki, K. Yamada, S. Hasegawa, D. Mizuki,M. Mizuki, I. Hatayama, K. Sekikawa, Y. Tagawa, et al. 2000. Roles of en-dogenous cytokines in liver apoptosis of mice in lethal Listeria monocytogenesinfection. FEMS Immunol. Med. Microbiol. 28: 335–341.

13. Lee, J., G. Cacalano, T. Camerato, K. Toy, M. W. Moore, and W. I. Wood. 1995.Chemokine binding and activities mediated by the mouse IL-8 receptor. J.Immunol. 155: 2158–2164.

14. Das, S. T., L. Rajagopalan, A. Guerrero-Plata, J. Sai, A. Richmond,R. P. Garofalo, and K. Rajarathnam. 2010. Monomeric and dimeric CXCL8 areboth essential for in vivo neutrophil recruitment. PLoS ONE 5: e11754. Availableat: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0011754.

15. Hatano, E., and D. A. Brenner. 2001. Akt protects mouse hepatocytes from TNF-alpha- and Fas-mediated apoptosis through NK-kappa B activation. Am. J.Physiol. Gastrointest. Liver Physiol. 281: G1357–G1368.

16. Jiwani, S., Y. Wang, G. C. Dowd, A. Gianfelice, P. Pichestapong, B. Gavicherla,N. Vanbennekom, and K. Ireton. 2012. Identification of components of the hosttype IA phosphoinositide 3-kinase pathway that promote internalization ofListeria monocytogenes. Infect. Immun. 80: 1252–1266.

17. Santos, S. A., D. R. Andrade, and D. R. Andrade Junior. 2005. Rat hepatocyteinvasion by Listeria monocytogenes and analysis of TNF-alpha role in apoptosis.Rev. Inst. Med. Trop. Sao Paulo 47: 73–80.

18. Jin, S., R. M. Ray, and L. R. Johnson. 2008. TNF-alpha/cycloheximide-inducedapoptosis in intestinal epithelial cells requires Rac1-regulated reactive oxygenspecies. Am. J. Physiol. Gastrointest. Liver Physiol. 294: G928–G937.

19. Kim, Y.-S., M. J. Morgan, S. Choksi, and Z.-G. Liu. 2007. TNF-induced acti-vation of the Nox1 NADPH oxidase and its role in the induction of necrotic celldeath. Mol. Cell 26: 675–687.

20. Schmeck, B., W. Beermann, V. van Laak, B. Opitz, A. C. Hocke,K. Meixenberger, J. Eitel, T. Chakraborty, G. Schmidt, H. Barth, et al. 2006.Listeria monocytogenes induced Rac1-dependent signal transduction in endo-thelial cells. Biochem. Pharmacol. 72: 1367–1374.

21. Modha, R., L. J. Campbell, D. Nietlispach, H. R. Buhecha, D. Owen, andH. R. Mott. 2008. The Rac1 polybasic region is required for interaction with itseffector PRK1. J. Biol. Chem. 283: 1492–1500.

22. Williams, C. L. 2003. The polybasic region of Ras and Rho family smallGTPases: a regulator of protein interactions and membrane association and a siteof nuclear localization signal sequences. Cell. Signal. 15: 1071–1080.

23. Casey, P. J., P. A. Solski, C. J. Der, and J. E. Buss. 1989. p21ras is modified bya farnesyl isoprenoid. Proc. Natl. Acad. Sci. USA 86: 8323–8327.

24. Kreck, M. L., D. J. Uhlinger, S. R. Tyagi, K. L. Inge, and J. D. Lambeth. 1994.Participation of the small molecular weight GTP-binding protein Rac1 in cell-free activation and assembly of the respiratory burst oxidase. Inhibition bya carboxyl-terminal Rac peptide. J. Biol. Chem. 269: 4161–4168.

25. Woodward, M. J., J. de Boer, S. Heidorn, M. Hubank, D. Kioussis, O. Williams,and H. J. M. Brady. 2010. Tnfaip8 is an essential gene for the regulation ofglucocorticoid-mediated apoptosis of thymocytes. Cell Death Differ. 17: 316–323.

26. Zhang, X., J. Wang, C. Fan, H. Li, H. Sun, S. Gong, Y. H. Chen, and Y. Shi.2009. Crystal structure of TIPE2 provides insights into immune homeostasis.Nat. Struct. Mol. Biol. 16: 89–90.

27. Choi, K.-C., S.-H. Kim, J.-Y. Ha, S.-T. Kim, and J. H. Son. 2010. A novel mTORactivating protein protects dopamine neurons against oxidative stress byrepressing autophagy related cell death. J. Neurochem. 112: 366–376.

28. Strimpakos, A. S., E. M. Karapanagiotou, M. W. Saif, and K. N. Syrigos. 2009.The role of mTOR in the management of solid tumors: an overview. CancerTreat. Rev. 35: 148–159.

29. Weichhart, T., G. Costantino, M. Poglitsch, M. Rosner, M. Zeyda,K. M. Stuhlmeier, T. Kolbe, T. M. Stulnig, W. H. Horl, M. Hengstschlager, et al.2008. The TSC-mTOR signaling pathway regulates the innate inflammatoryresponse. Immunity 29: 565–577.

30. Saci, A., L. C. Cantley, and C. L. Carpenter. 2011. Rac1 regulates the activity ofmTORC1 and mTORC2 and controls cellular size. Mol. Cell 42: 50–61.

31. Pamer, E. G. 2004. Immune responses to Listeria monocytogenes. Nat. Rev.Immunol. 4: 812–823.

32. Rogers, H. W., M. P. Callery, B. Deck, and E. R. Unanue. 1996. Listeria mon-ocytogenes induces apoptosis of infected hepatocytes. J. Immunol. 156: 679–684.

33. Zheng, S.-J., J. Jiang, H. Shen, and Y. H. Chen. 2004. Reduced apoptosis andameliorated listeriosis in TRAIL-null mice. J. Immunol. 173: 5652–5658.

34. Gurung, P., D. Rai, S. A. Condotta, J. C. Babcock, V. P. Badovinac, andT. S. Griffith. 2011. Immune unresponsiveness to secondary heterologous bac-terial infection after sepsis induction is TRAIL dependent. J. Immunol. 187:2148–2154.

35. dos Santos, S. A., D. R. de Andrade Junior, and D. R. de Andrade. 2011. TNF-aproduction and apoptosis in hepatocytes after Listeria monocytogenes and Sal-monella Typhimurium invasion. Rev. Inst. Med. Trop. Sao Paulo 53: 107–112.

36. Harty, J. T., and M. J. Bevan. 1995. Specific immunity to Listeria monocytogenesin the absence of IFN gamma. Immunity 3: 109–117.

37. Jensen, E. R., A. A. Glass, W. R. Clark, E. J. Wing, J. F. Miller, andS. H. Gregory. 1998. Fas (CD95)-dependent cell-mediated immunity to Listeriamonocytogenes. Infect. Immun. 66: 4143–4150.

38. Ben Moshe, T., H. Barash, T.-B. Kang, J.-C. Kim, A. Kovalenko, E. Gross,M. Schuchmann, R. Abramovitch, E. Galun, and D. Wallach. 2007. Role ofcaspase-8 in hepatocyte response to infection and injury in mice. Hepatology 45:1014–1024.

39. Conlan, J. W., P. L. Dunn, and R. J. North. 1993. Leukocyte-mediated lysis ofinfected hepatocytes during listeriosis occurs in mice depleted of NK cells orCD4+ CD8+ Thy1.2+ T cells. Infect. Immun. 61: 2703–2707.

40. Haschtmann, D., H. J. Gerber, and M. E. Mielke. 2005. Cytotoxic activity ofmurine resident peritoneal cells against Listeria monocytogenes-infected hepa-tocytes in vitro. Microbes Infect. 7: 1177–1183.

41. Gouin, E., M. Adib-Conquy, D. Balestrino, M.-A. Nahori, V. Villiers, F. Colland,S. Dramsi, O. Dussurget, and P. Cossart. 2010. The Listeria monocytogenes InlCprotein interferes with innate immune responses by targeting the IkappaB kinasesubunit IKKalpha. Proc. Natl. Acad. Sci. USA 107: 17333–17338.

42. da Silva, C. V., L. Cruz, Nda. S. Araujo, M. B. Angeloni, B. B. Fonseca,Ade. O. Gomes, Fdos. R. Carvalho, A. L. R. Goncalves, and Bde. F. Barbosa.2012. A glance at Listeria and Salmonella cell invasion: different strategies topromote host actin polymerization. Int. J. Med. Microbiol. 302: 19–32.

43. Veiga, E., and P. Cossart. 2005. Listeria hijacks the clathrin-dependent endocyticmachinery to invade mammalian cells. Nat. Cell Biol. 7: 894–900.

44. You, Z., H. Ouyang, D. Lopatin, P. J. Polver, and C. Y. Wang. 2001. Nuclearfactor-kappa B-inducible death effector domain-containing protein suppressestumor necrosis factor-mediated apoptosis by inhibiting caspase-8 activity. J.Biol. Chem. 276: 26398–26404.

8 TNFAIP8 IN APOPTOSIS AND LISTERIOSIS

by guest on February 16, 2018http://w

ww

.jimm

unol.org/D

ownloaded from