of March 27, 2018.This information is current as

T Cell Death in Persistent InfectionBid and Bim Collaborate during Induction of

Strasser and Gabrielle T. BelzFrederick Masson, Fiona Kupresanin, Adele Mount, Andreas

ol.1001918http://www.jimmunol.org/content/early/2011/02/21/jimmun

published online 21 February 2011J Immunol 

MaterialSupplementary

8.DC1http://www.jimmunol.org/content/suppl/2011/02/22/jimmunol.100191

        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 © 2011 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 March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Bid and Bim Collaborate during Induction of T Cell Death inPersistent Infection

Frederick Masson, Fiona Kupresanin, Adele Mount, Andreas Strasser,

and Gabrielle T. Belz

Upon Ag encounter, naive T cells undergo extensive Ag-driven proliferation and can differentiate into effector cells. Up to 95% of

these cells die leaving a small residual population of T cells that provide protective memory. In this study, we investigated the

contribution of the BH3-only family protein Bid in the shutdown of T cell responses after acute and persistent infection. Influenza

virus pathogenicity has been proposed to be mediated by a peptide encoded in the basic polymerase (PB1-RF2) acting through

Bid. In our experiments, we found that after acute infection with influenza virus, mice lacking Bid had normal expansion and

contraction of Ag-specific CD8+ T cells. However, in chronic g-herpesvirus infection, Bid-deficient virus-specific CD8+ T cells

expanded normally but failed to contract fully and were maintained at ∼2-fold higher levels. Previously, we have demonstrated

that Bim plays a prominent role in T cell shutdown in persistent infection by cooperating with the death receptor Fas, which

regulates apoptosis in response to repeated TCR signaling. Bid lies at the nexus of these two signaling pathways, thus we reasoned

that Bid and Bim might cooperate in regulation of T cell shutdown in persistent infection. In this study, we observed that the

combined loss of Bid and Bim synergistically enhanced the persistence of CD8+ T cells during g-herpesvirus infection. Thus, these

data uncover a role for Bid in coordinating apoptotic signaling pathways to ensure appropriate shutdown of T cell immune

responses in the setting of persistent Ag exposure. The Journal of Immunology, 2011, 186: 000–000.

The mechanisms that underpin regulation of the balancebetween activation of immune cells to maximize reactivityto foreign Ags while minimizing responses to self are

central to our understanding of immunity. Apoptosis plays a criticalrole in the development of immune cells and in the maintenance oftissue homeostasis. This is achieved through the regulation ofapoptotic family members that shape the immune repertoire andterminate immune responses (1). Whether a cell lives or dies islargely due to the interactions between antiapoptotic and proa-poptotic members of the B cell lymphoma 2 (Bcl-2) family. Theantiapoptotic proteins (Bcl-2, Bcl-XL, Bcl-W, A1, and Mcl-1)share sequence homology in their Bcl-2 homology (BH) do-mains. The proapoptotic members are divided into two groups: themultidomain proteins (e.g., Bax and Bak) and those that sharesimilarity in the BH3 region (e.g., Bim/Bod, Bid, Bad, Noxa, Hrk/

DP5, Bik, Blk/NBk, Puma/Bbc3). This latter group of BH3-onlyproteins can bind to the prosurvival Bcl-2 family members therebyremoving the inhibition of Bax and Bak and inducing apoptosis bypermeabilizing the outer mitochondrial membrane (2, 3).In mammals, two pathways of apoptosis, the “Bcl-2–regulated”

(also called “intrinsic” or “mitochondrial”) and the “death re-ceptor”-induced (also called “extrinsic”) synergistically regulatethe shutdown of CD8+ T cell responses after a chronic immuneresponse (4, 5). The relative contributions of each of these path-ways vary depending on the nature of the response—acute versuschronic. In an acute immune response, cessation of lymphocyterecruitment and proliferation is thought to be triggered by a de-cline in cytokine levels and is mediated (largely) through activa-tion of the BH3-only proteins Bcl-2 interacting mediator of celldeath (Bim) and p53 upregulated modulator of apoptosis (Puma)(6–9). Bim and Puma, in contrast to other proapoptotic molecules,bind to all prosurvival Bcl-2–like proteins with high affinity andtherefore play particularly prominent roles in the initiation ofapoptosis (10, 11). In chronic immune responses elicited by per-sistent pathogens or tissue self-antigens, the “death receptor” (Fasligand [FasL]) and the “Bcl-2–regulated” (Bim) pathways col-laborate in effecting the removal of Ag-activated CD8+ T cells(12–14). In this setting, the death receptor pathway is thought tobe triggered by repeated TCR ligation-mediated induction ofFasL, leading to autocrine and/or paracrine Fas stimulation,whereas Bim may be induced in response to a decline in cytokinelevels or repeated TCR stimulation-induced calcium flux (5, 15,16). More recently, Bim, FasL, and BH3-interacting domain deathagonist (Bid) have been shown to regulate collaboratively pro-tection against fatal autoimmune hepatitis; however, a similarpotential synergistic role in persistent infection has not yet beenexamined (17).Bid lies at the nexus of the extrinsic and intrinsic pathways:

ligation of Fas results in caspase-8–mediated cleavage of Bid to t-Bid and its subsequent translocation to the mitochondria where the

Walter and Eliza Hall Institute of Medical Research, Melbourne, Victoria 3052,Australia

Received for publication June 10, 2010. Accepted for publication January 19, 2011.

This work was supported by the National Health and Research Council of Australia(to G.T.B. and A.S.), the Leukemia and Lymphoma Society (to A.S.), the JuvenileDiabetes Research Foundation (to A.S.), a Swiss National Science Foundation fel-lowship (to F.M.), a University of Melbourne Research Scholarship (to A.M.), and bya Viertel Fellowship and a Howard Hughes Medical Institute International Fellowship(to G.T.B.).

Address correspondence and reprint requests to Dr. Gabrielle T. Belz, Walter andEliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3052,Australia. E-mail address: belz@wehi.edu.au

The online version of this article contains supplemental material.

Abbreviations used in this article: Bcl-2, B cell lymphoma 2; BH, Bcl-2 homology;Bid, BH3-interacting domain death agonist; Bim, Bcl-2 interacting mediator of celldeath; DC, dendritic cell; FasL, Fas ligand; g-HV, g-herpesvirus; LN, lymph node;MHV-OVA, murine g-herpesvirus-68 expressing ovalbumin; NP, nucleoprotein ofinfluenza virus; PA, acid polymerase of influenza virus; PB1-RF2, basic polymerase1–reading frame 2; PI, propidium iodide; PR8, influenza A/PR/8/34; Puma, p53upregulated modulator of apoptosis; WT, wild-type; X31, influenza HKx31 (H3N2)virus.

Copyright� 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1001918

Published February 21, 2011, doi:10.4049/jimmunol.1001918 by guest on M

arch 27, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

intrinsic death pathway is engaged (18). Both acute and persistentpathogens have developed a number of strategies to co-opt com-ponents of the apoptotic machinery to impair cell death and totarget components of the Bid pathway. Influenza A/PR/8/34 (PR8)virus is of particular interest as it encodes the basic polymerase 1(basic polymerase 1-reading frame 2; PB1-RF2) (19) that is hy-pothesized to induce cell death by activating the Bcl-2–regulatedapoptotic pathway by sensitizing mitochondria to the proapoptoticeffects of t-Bid (20). This mechanism is thought to be criticalin defining influenza virus pathogenicity. Previously, we did notdetect an effect of loss of Bid on the immune response to acuteHSV-1 infection (8); however, it is important to determine whetherthe specific action of PB1-RF2 in influenza pathogenicity is me-diated through Bid in vivo or by an alternate pathway. In additionto the purported action of influenza on Bid, persistent viruses suchas herpes and pox viruses express viral FLIPs, which inhibit theactivation of caspase-8, the enzyme that cleaves and activates Bid(21). We therefore reasoned that Bid, similar to Bim, might reg-ulate T cell expansion or contraction after acute and persistentinfection either in a T cell intrinsic manner or by modulating thesurvival of APCs. In addition, given the synergistic interactionsbetween Bim and FasL and the importance of Bid in linking theseintrinsic and extrinsic pathways particularly in persistent in-fection, Bid might act with Bim to optimize T cell shutdown.

Materials and MethodsMice

All mice were bred and maintained in specific pathogen-free conditions atthe Walter and Eliza Hall Institute of Medical Research. Bid-deficient micewere generated on an inbred C57BL/6 background using C57BL/6-derivedembryonic stem cells (22); bim2/2 (23) and bim2/2bid2/2 (17) have beenpreviously described; transgenic strains Ly5.2 OT-I (Ly5.2.OT-I; H-2Kb–restricted OVA257–264 specific TCR) and Ly5.1 3 Ly5.2.F1 bid2/2OT-I.C57BL/6 and Ly5.1 (B6.SJL-PtprcaPep3b/BoyJ) mice were generated andmaintained at the Walter and Eliza Hall Institute of Medical Research.Mice were used between 6 and 12 wk of age, and all experiments wereperformed according to the guidelines of the Walter and Eliza Hall In-stitute of Medical Research Animal Ethics Committee.

Virus infections

Mice were anesthetized with methoxyflurane and then infected intranasallywith 3 3 104 PFU recombinant murine g-herpesvirus-68 expressing ov-albumin (MHV-OVA) (24) or with 104.5 PFU HKx31 (H3N2; X31) in-fluenza virus (25, 26) or 101.39 PFU PR8 influenza virus.

CD8+ T cell enrichment and CFSE labeling

CD8+ T cells were enriched from spleen and lymph nodes (LNs) bygenerating a single-cell suspension and incubating the cells in a mixture ofoptimally titrated Abs against Mac-1 (M1/70), Mac-3 (F4/80), Ter-119,GR1 (RB6-8C5), MHC class II (M5/114), and CD4 (GK1.5) for 30 min onice. The Ab-bound cells were removed using anti-rat IgG Ab conjugatedmagnetic beads (Dynabeads; Dynal). OT-I transgenic cells contained 87–96% TCR Va2+ CD8+ cells.

In vitro-generated effector OT-I CD8+ T cells were labeled with CFSE (5mM; Molecular Probes) by incubating 107 purified cells per milliliter with5 mM CFSE for 10 min at 37˚C as previously described (27). CFSE-labeled effector OT-I CD8+ T cells (2.5 3 106–5 3 106) were adop-tively transferred by tail vein injection in 200 ml PBS into recipient mice.

Cell surface staining and FACS analysis

To block nonspecific binding of Abs, cells were incubated in rat IgG (1 mg/ml) together with anti-FcgR mAb (CD16/32, clone 2-4G21) for 10 min onice. Samples were then stained with fluorochrome-labeled Abs againstCD11c (HL3), Va2 (B20.1), CD8a (53-6.7), CD62L (Mel-14), KLRG1(2FY), IL-7R (A7R34), Ly5.1 (A20-1.1), and Ly5.2 (104) (all fromeBioscience or BD Pharmingen). Virus-specific CD8+ T cells were enumer-ated by staining with PE-coupled tetrameric H-2b MHC class I complexesloaded with epitopes of MHV-OVA virus (OVA, OVA257–264 SIINFEKL,H-2Kb–restricted; p79 ORF61524–531, TSINFVKI, H-2K

b–restricted; p56ORF6487–495, AGPHNDMEI, H-2Db–restricted) and influenza virus nu-

cleoprotein (NP; DbNP366–374, H-2Db–restricted) and influenza virus acidic

polymerase (PA; DbPA224–233, H-2Db–restricted). Viable cells were ana-

lyzed by flow cytometry using propidium iodide (PI) exclusion.

Intracellular cytokine staining

To measure Ag-specific cytokine expression, CD8+ T cell-enrichedsplenocytes were stimulated with 1 mM OVA (OVA257–264) or p79 (p79ORF61524–531) peptide in presence of Golgi Stop (BD Pharmingen) for 5 h.Cells were then stained for surface molecules CD8a (53-6.7), Ly5.1 (A20-1.1), and Ly5.2 (104), fixed and permeabilized using Cytofix/Cytopermreagent (BD Biosciences), and then stained for intracellular IFN-g (anti-XMG 1.2-PE), TNF-a (MP6-XT22), and washed and analyzed by flowcytometry.

Lymphocyte isolation from the lung

Mononuclear cells were purified from the lung by generating single-cellsuspensions, followed by centrifugation on Histopaque density gradient(1.077 g/ml; Sigma-Aldrich) for 18 min at 2000 rpm at room temperature.Lymphocytes from bronchoalveolar lavage were isolated by washing thelung three times with HEPES-buffered Eagle’s medium supplemented with2.5% FCS.

Dendritic cell enrichment

Dendritic cells (DCs) were purified from the spleen as previously described(28).

In vitro generation of effector CD8+ T cells

Effector OT-I T cells were generated by culturing OT-I T cells with OVA-pulsed DCs (1 mg/ml OVA peptide) in complete RPMI 1640 medium(mouse tonicity RPMI 1640 containing 10% FCS, 50 mM 2-mercaptoe-thanol, 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml strepto-mycin) supplemented with 30 U/ml human recombinant IL-2. DCs werecultured with T cells at a ratio of 1:10. Cells were placed into fresh me-dium containing IL-2 after 3 d. On day 7 after initiation of the culture, cellswere collected and subjected to a second round of stimulation with OVApeptide-pulsed DCs and then cultured for an additional 3 d prior toadoptive transfer.

Flow cytofluorometric assay for T cell-mediated cytotoxicity

OVA-specific CTLs were generated in vitro. DCs enriched from the spleensof wild-type (WT) or bid2/2 mice were pulsed with 100 ng/ml OVApeptide for 1 h at 37˚C. Graded numbers of OT-I CTLs were incubated in96-well round-bottom plates before the addition of 5 3 104 OVA peptide-pulsed or unpulsed DC targets. After 3-h incubation at 37˚C, cells werewashed and stained with mAbs against TCRVa2 (B20.1) and CD11c(HL3) and washed before staining with FITC-conjugated annexin V Ab(BD Pharmingen) and analysis by flow cytometry. The percentage ofkilling was then calculated as follows: [(percentage of annexin+ DCs in-cubated with T cells 2 percentage of annexin+ DCs cultured alone)/(percentage of annexin+ DCs cultured alone)] 3 100.

ResultsPB1-RF2–mediated pathogenicity is independent of Bid

Influenza infection of APCs is critical to induction of the T cellresponse. We hypothesized that if Bid is a major mediator of in-fluenza virus pathogenicity (20) through the action of PB1-RF2, itwould act through APCs as it does not efficiently infect T cells.Modification of Ag presentation by induction of death of APCsthrough the action of PB1-RF2 targeting the apoptotic pathwayshould result in differences in induction of CD8+ T cell responses.To test whether loss of Bid in host cells enhanced CD8+ T cellresponses and protection against PB1-RF2–mediated cell death,we inoculated bid2/2 and WT mice with one of two strains ofinfluenza A virus, namely X31 or the parent strain, PR8. X31 isa recombinant influenza virus in which the non-coating genes(nucleoprotein NP, acid polymerase PA, basic polymerase PB1,PB2, non-structural NS1, NS2, matrix M1, M2) are derived fromthe PR8 virus and thus are completely homologous but can differin pathogenicity. In C57BL/6 mice, intranasal infection with X31

2 Bid AND PERSISTENT INFECTION

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

generates a relatively mild respiratory infection compared withthat of PR8, and it is speculated that the action of PB1-RF2 maybe of greater importance in more highly pathogenic influenzainfections. Thus, first we tracked whether loss of Bid enhancedrecovery or protection after influenza infection. Primary infection

with either of these influenza viruses resulted in similar weightloss in bid2/2 and WT mice (Fig. 1A) and no differences in lunghistopathology (data not shown). In the more virulent PR8 in-fection (Fig. 1A, right panel), no statistical differences were ob-served in the number or timing of death of fatally affected mice.

FIGURE 1. Bid loss has no impact on CD8+

T cell responses to acute influenza X31 or PR8

viral infection. Bid2/2 or WT mice were

infected intranasally with X31 (A [left panel], B)

or PR8 (A [right panel], D) influenza viruses. A,

Body weight of influenza-infected bid2/2 (black

symbols) or WT (+/+, white symbols) mice was

measured daily for up to 10 d postinfection.

Data show the percentage initial body weight

(mean6 SEM) of 9–12 mice for each genotype.

Mice that were killed or that died during PR8

infection are shown as individual symbols. B–D,

Enumeration of influenza-specific (DbNP366 and

DbPA224) CD8+ T cells within the spleen (B, D),

mediastinal LN and lung (C, D) at days 8–10

(expansion phase of response) and 21 d (mem-

ory, contraction phase) after primary challenge

with influenza virus. Data show the mean 6SEM of the number of virus-specific CD8+

T cells from 6 to 10 mice at each time point

pooled from two experiments. For lung, nine

individual animals were analyzed, and medias-

tinal LNs were pooled from four to five mice

from each of two experiments; individual values

for pooled mediastinal LNs in each experiment

are shown as circles. Flow cytometric plots show

representative staining of PE-conjugated MHC

class I tetramers for DbNP366 or DbPA224 on

CD8+ T cell-enriched splenocytes from WT and

bid2/2 mice 21 d postinfection with X31 (B,

right panels).

FIGURE 2. Loss of Bid enhances

accumulation of virus-specific CD8+

T cell populations in bid2/2 mice after

persistent g-HV infection. Bid2/2 and

WT (+/+) mice were infected intrana-

sally with recombinant MHV-OVA. A,

The number of virus-specific CD8+ T

cells were enumerated by staining with

MHC class I/peptide tetramers. Data

show the mean 6 SEM of the number

of virus-specific CD8+ T cells for 14,

72, and 120 d postinfection pooled

from two to three experiments and rep-

resent 8–10 mice for each time point.

Statistical differences were assessed by

a two-tailed Student t test: *p , 0.05,

**p , 0.01. B, Flow cytometric plots

show representative KbOVA257, Dbp56,

and Kbp79 tetramer staining of CD8+ T

lymphocyte-enriched splenocytes from

WT or bid2/2 mice 14 d after infection

with MHV-OVA.

The Journal of Immunology 3

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Thus, the loss of Bid did not protect against the more virulent PR8influenza infection.We then analyzed whether Bid influenced the generation or loss

of influenza-specific CD8+ T cell responses. The absence of Biddid not result in differences in the number or frequency of DbNPor DbPA-specific CD8+ T cells between bid2/2 and WT mice atthe peak of the response (day 10) (Fig. 1B, X31; 1D, PR8) or at thetime of established immunological memory (day 21) in the spleen,lung, or mediastinal LNs for either infection (Fig. 1B–D). Similarresults were obtained in mixed bone marrow chimeric mice inwhich bid2/2 and WT cells exposed to similar Ag loads and cy-tokine milieu can be analyzed simultaneously for cell intrinsicdifferences during infection (data not shown). Thus, althoughPB1-RF2 can exert increased pathogenicity in some infectioussettings as highlighted by mutation of this sequence (6, 7, 20, 29–31), it does not appear to do this in vivo through a Bid-dependentpathway in mice as was initially predicted.

Loss of Bid causes accumulation of virus-specific CD8+ T cellsduring chronic infections

Next we asked whether Bid plays a role in persistent infection. Toidentify whether this might be the case, bid2/2 or WT mice wereinfected with recombinant MHV-OVA by intranasal inoculation.This virus generates an acute lytic infection in the lung andmediastinal LNs after which it disseminates to the spleen andestablishes persistent reservoirs (32, 33). After 2 wk, the peak oflytic infection, there was no significant alteration in the total cellnumber, the activation status, or function (Supplemental Fig. 1) ofvirus-specific CD8+ T cells in the spleens of bid2/2 infected micecompared with that in WT mice (Fig. 2). Although loss of Bid atthis early stage of infection did not cause any obvious changein lymphocyte numbers or viral titers, we further analyzed theimmune response at later stages of MHV-OVA infection (days 72and 120), when persistent exposure to Ag can drive changes inT cell homeostasis (Fig. 2) (13). The virus-specific CD8+ T cellresponse to three different epitopes, the surrogate engineered vi-ral epitope KbOVA and two endogenous viral epitopes, Dbp56(AGPHNDMEI from ORF6487–495) and Kbp79 (TSINFVKI fromORF61524–531), which are associated with lytic and latent geneexpression, respectively, were measured. Four months post-infection, bid2/2 mice accumulated 2-fold more KbOVA and Db

p56-specific CD8+ T cell populations in the spleen compared withthat in WT mice. A consistently expanded population of Kbp79cells was also observed, although this was not as marked as thatfor KbOVA and Dbp56-specific CD8+ T cell populations perhapsreflecting differences in overall expression of the different Agsduring persistent infection. A similar trend was observed in thelungs and the mediastinal LNs toward the KbOVA and Dbp56epitopes, although this did not reach statistical significance(Supplemental Fig. 2). In this setting, T cells lacking Bid main-tained their functional capacity to produce cytokines such as IFN-g and TNF-a after short-term in vitro restimulation (SupplementalFig. 4). This is consistent with our previous observations thatT cells exposed to persistent g-herpesvirus (g-HV) maintainfunction (8, 13, 34). Overall, these data suggest that Bid can playa role in the regulation of virus-specific CD8+ T cell homeostasisduring persistent infection.

Virus-specific CD8+ T cell accumulation in bid2/2 is notdriven by changes in DC life span

BH3-only proteins have broad effects on the survival of immunecells. One explanation for the accumulation of bid2/2 T cells isthat they may be driven to increased proliferation secondary toenhanced survival of DCs in the absence of Bid. To determine

whether differences in DC survival may be important, we isolatedDCs from naive WT mice and mice deficient in BH3-only proteinsBid and Bim and followed their survival in vitro (Fig. 3A). We didnot observe any significant differences in survival between bid2/2

and WT DCs, but bim2/2 DCs did display a survival advantage(Fig. 3A). This analysis, however, does not exclude the possibilitythat Bid may play a role in vivo when activated Ag-loaded DCsencounter primed and differentiated effector CD8+ T cells that killDCs expressing cognate Ags. Indeed, granzyme B released byeffector CD8+ T cells can cleave Bid in target cells, thereby ac-tivating the Bid-dependent apoptotic pathway (35, 36). Thus, inthe absence of Bid activation, DCs may be rendered more resistantto granzyme/perforin-mediated T cell-induced cytolysis. To in-vestigate this, we established a cytofluorometric assay that allowsthe detection of apoptosis in small numbers of target cells (37).Purified DCs from bid2/2 and WT mice pulsed with OVA257–264

peptide were killed with equal efficiency by in vitro-activatedOVA-specific CD8+ T cells suggesting that Bid deficiency doesnot protect DCs from T cell-mediated killing in this setting (Fig.3B). Furthermore, when DCs isolated from mice 14 d after MHV-

FIGURE 3. Loss of Bid does not affect survival of APCs in vitro. APCs

were enriched from spleens of bid2/2, bim2/2, and WT mice by DNase/

collagenase digestion and density gradient centrifugation. A, APCs were

cultured in complete medium for up to 3 d, and the viability of CD11c+

cells was determined at the indicated time points by staining with anti-

CD11c mAb, FITC-conjugated annexin V, and PI. Data show the mean 6SEM percentage of annexin2 PI2 CD11c+ cells pooled from two in-

dependent experiments each performed in triplicate. B, APCs were pulsed

for 1 h with 100 ng/ml OVA257–264 peptide or left unpulsed, then cocul-

tured with activated OVA-specific CD8+ T cells at various ratios of APCs/

OVA-specific T cell. OVA-specific killing was determined by staining with

anti-CD11c, FITC-conjugated annexin V, and PI and analysis by flow

cytometry. OVA-specific killing was calculated as follows: [(percentage

annexin+ CD11c+ DCs with OVA-specific T cells 2 percentage annexin+

CD11c+ DCs without T cells) / (percentage annexin+ CD11c+ DCs without

T cells)]3 100. Data represent the mean6 SEM of the percentage specific

killing at the indicated OT-I/APC ratio. C, Analysis of annexin V staining

on DCs isolated from mice 14 d after infection with MHV-OVA. Data

show the mean and SEM of four to five mice of each genotype.

4 Bid AND PERSISTENT INFECTION

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

OVA infection were examined, they showed equivalent levels of

annexin V staining suggesting that infection itself did not alter the

potential effects of loss of Bid (Fig. 3C). Thus, collectively these

data demonstrate that Bid does not contribute to the death of

splenic DCs.

Bid intrinsically regulates the death of virus-specific T cellsundergoing repeated antigenic stimulation

Given the above findings, it seemed likely that the key effects ofBid deficiency in T cells were intrinsic and mediated by increasedresistance to cell death. To analyze specifically the cell intrinsic

FIGURE 4. Loss of Bid affects survival but not proliferation of effector CD8+ T cells after stimulation with viral Ag. A, Schematic representation of the

experimental approach for adoptive transfer of CD8+ T cells. i.n., intranasal. B, 2.53 106 CFSE-labeled naive or effector WT (CD45.2) or bid2/2 (CD45.1/

2) OT-I CD8+ T cells were adoptively transferred into CD45.1+ naive or MHV-OVA (3 d) recipients. Three days (6 d infection; B [upper panels]) or 12 d (15

d infection; B [lower panels]) after transfer, the mediastinal LN, lung, and spleen were removed and dissociated, and cells were stained with anti-CD8a–

allophycocyanin and anti-CD45.1–PE Abs to analyze proliferation of T cells. The percentage of CD45.1/CD45.2 bid2/2 or CD45.2 WT TCRVa2+CD8+

OVA-specific T cells recovered from different organs are indicated on dot plots. Histograms show the proliferation of bid2/2 (black line) or WT (red)

TCRVa2+ OT-I CD8+ T cells. C, Enumeration of bid2/2 and WT TCRVa2+CD8+ OT-I T cells in spleen, mediastinal LN, and lung after MHV-OVA

infection 12 d after adoptive transfer of transgenic T cells. Data show the mean 6 SEM of seven to eight individual mice pooled from two experiments.

Mediastinal LNs were analyzed as individuals or pooled from three to four animals/experiments. Individual results are indicated by circles. Statistically

significant differences, as assessed by a two-tailed paired Student t test, are indicated: *p , 0.05, **p , 0.01.

The Journal of Immunology 5

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

effects of the loss of Bid on virus-specific CD8+ T cells after thepeak of the response, we generated effector CD8+ T cells in vitroby activating OT-I CD8+ T cells using OVA-peptide pulsed DCs.Effector bid2/2OT-I (Ly.5.1/Ly5.2) and WT OT-I CD8+ (Ly5.2)T cells (.95% cells were CD62L2) were labeled with CFSE,mixed in a 1:1 ratio, and adoptively transferred into either naiveLy5.1 mice or mice infected 3 d earlier with MHV-OVAvirus (Fig.4A). This approach allowed parallel tracking of CD8+ T cells thateither lacked or retained expression of Bid but that expresseda TCR of equivalent affinity and avidity. Three days after transfer,flow cytometric analysis of the dilution of CFSE showed that bothbid2/2OT-I and WT OT-I CD8+ T cells proliferated similarly inthe mediastinal LN and lungs of mice (Fig. 4B, top panels). Asexpected at this early time point (6 d postinfection), proliferationwas most prominent in the mediastinal LN (the site of immuneactivation) while few dividing cells were detected in the spleen ofinfected mice. However, at 12 d after transfer, at least 5-fold morebid2/2 OT-I CD8+ T cells were detected in spleen and 3-fold morein the lungs compared with WT OT-I cells (Fig. 4B [lower panels],4C; Supplemental Fig. 3). By contrast, there was no differencein the number of bid2/2OT-I CD8+ T cells and WT OT-I CD8+

T cells in the spleens of naive mice over the same time frameindicating that both genotypes survived equally well at steadystate (Fig. 4B, 4C). Thus, Bid regulates the death of effector virus-specific CD8+ T when exposed to chronic Ag stimulation.

Bid cooperates with Bim to regulate virus-specific CD8+ T celldeath during persistent virus infection

Although in the absence of Bid more CD8+ T cells accumulate inpersistent infection, this failure to shut down the T cell response ismore modest than what we have previously shown for Bim (13).However, BH3-only proteins can act synergistically to ensure fine-tuning of the T cell shut-down process. For example, Bim and Fascooperate to regulate T cell shutdown during persistent infection(13). In addition, Bim, Bid, and Fas work collaboratively in reg-ulating T cell shutdown in hepatitis that may occur under anti-genic stimulation (17). As Bid connects the death receptor and theBcl-2–regulated apoptotic pathways, we asked whether the modestaccumulation of bid2/2 virus-specific T cells during repeatedexposure to Ag was mediated by Bim and Fas dependence on Bid.If Bid operated to optimize CD8+ T cell responses, and thuscomplement the action of Bim, we hypothesized that concurrentloss of Bid and Bim would result in enhanced survival of T cellsover loss of either Bid or Bim alone. Conversely, if Bid does notplay a role, loss of both BH3-only family members would notresult in any survival advantage to CD8+ T cells compared withloss of Bim. To address this issue, bid2/2, bim2/2, and bim2/2

bid2/2 double-deficient mice were infected intranasally with MHV-OVA virus, and the virus-specific CD8+ T cell immune responsewas analyzed 40 d later. Although the number of splenocytesin bim2/2bid2/2 mice was significantly higher in bim2/2 mice,the number of CD8+ T cells was similar in both strains of mice(data not shown). Despite this, bim2/2bid2/2 mice had a 2-foldincrease of OVA-specific T cell numbers compared with those ofbim2/2 mice and was statistically significant (p , 0.05) (Fig. 5A).The number of p79 specific CD8+ T cells was similar in bothstrains of mice consistent with our earlier findings in bid2/2 miceperhaps suggesting the earlier elimination of infected cells by theincreased OVA-specific CD8+ T cells (Figs. 2, 5). Phenotypically,virus-specific CD8+ T cells isolated from WT, bid2/2, bim2/2, andbim2/2bid2/2 mice were similar, with comparable proportions ofshort-lived effector cells (KLRG1highIL-7Rlow) and memory pre-cursor cells found in spleen (KLRG1low/intIL-7Rhigh) (Fig. 5B).

Collectively, these data show that an unanticipated synergy existsbetween Bim and Bid in T cells responding to persistent Ag.

DiscussionIn this study, we demonstrated that Bid does not contribute to deathof Ag-specific CD8+ T cells during acute influenza infection. Thisis concordant with our previous observations using acute HSV-1infection in bid2/2 and WT mice (8). Similar to the scenarioobserved in influenza and HSV-1 infection (8), the induction ofg-HV–specific CD8+ T cells showed a pattern equivalent to that ofWT mice. However, when we extended the analysis to examinelate time points, over which time Ag is continuously expresseddue to persistent infection, we uncovered a role for Bid in pre-venting the accumulation of virus-specific CD8+ T cells. Loss ofBid resulted in approximately a 2-fold increase in the number ofvirus-specific CD8+ T cells detected when activated effector CD8+

T cells predominate in persistent infection. To delineate furtherthis phenomenon, activated virus-specific CD8+ T cells wereadoptively transferred into infected mice. Bid2/2 effector T cellsaccumulated preferentially compared with WT effector CD8+

T cells in response to Ag exposure. Ag appears to be a key factortriggering activation of the Bid-dependent pathway, as after acuteviral infections such as influenza or HSV (8), Bid-deficient cellsdo not gain a survival advantage over WT cells. In persistentg-HV infection, CD8+ T cell populations that recognized proteinsexpressed in lytic infection (active viral replication) showed en-hanced survival reflecting the role of Bid in regulating survival ofeffector-type CD8+ T cells exposed to persistent Ag. Intriguingly,in addition to the increased frequency of KLRG1+CD1272 T cellstypical of a classical effector cell type, we observed a significant

FIGURE 5. Bid and Bim cooperate to regulate the accumulation of

virus-specific CD8+ T cells during persistent viral infection. A, Bim2/2

bid2/2, bim2/2, bid2/2, and WT (+/+) mice were infected intranasally

with MHV-OVA and analyzed at 40 d postinfection. The total number of

virus-specific CD8+ T cells in spleen was determined by staining with PE-

conjugated MHC class I/peptide tetramers loaded with KbOVA257 and Kb

p79 (ORF61524) peptides. Data show the mean 6 SEM of the number of

virus-specific CD8+ T cells pooled from two experiments with seven to

nine animals analyzed for each genotype. Statistically significant differ-

ences were determined using a two-tailed Student t test: *p , 0.05. B,

Expression of activation markers KLRG1 and IL-7R on p79 and OVA-

specific CD8+ T cells isolated from spleen 40 d postinfection. Plots are

gated on CD8a+PI2 cells. Data are representative of two experiments, and

plots show the mean 6 SEM percentage for each quadrant.

6 Bid AND PERSISTENT INFECTION

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

population of cells that coexpressed KLRG1 and CD127. Thenature of these cells is currently under investigation, but theycould represent T cells that have only undergone partial activationand thus not down-modulated CD127 although we find no evi-dence for functional impairment (IFN-g and TNF-a secretion)(Supplemental Fig. 1, Supplemental Fig. 4). In vitro repeatedactivation of virus-specific CD8+ T cells prior to transfer into micechronically infected with g-HV mimicked the accumulation ofbid2/2 cells seen in endogenous T cell populations. Such an effectwas not seen when effector CD8+ T cells were adoptively trans-ferred into naive hosts. Although prolonged life span of APCscould perhaps account for such an effect, loss of Bid did not ap-pear to alter survival of splenic DCs. Thus, the contribution of Bidto T cell survival in our studies reflects a T cell intrinsic effect.This cell intrinsic effect of Bid is reinforced by the outcome ofadoptively transferred cells into WT infected hosts. It is strikingthat the phenotype of bid-deficient and bid/bim-deficient mice isless pronounced than that of the bim/fas double knockout mice(Bcl2l112/2Faslpr/lpr) (13). In this latter study, delineation of theindependent intrinsic effects of the compound loss of Bim and Fason T cells versus APCs was not assessed. Stranges et al. (38)demonstrated that Fas-dependent elimination of APCs is a majormechanism curbing immune responses in chronically activatedT cells, particularly in the setting of autoimmunity . As such,prolonged Ag presentation in the absence of both bim and fasis likely to magnify significantly the amplification of T cellsresponding to infection where systemic loss of these regulators,rather than T cell-specific deletion, has been induced. This mayaccount for the greater difference observed in the bim/fas-deficientmice, where both bim (Fig. 3) and fas (38) contribute to survival ofDCs, compared with the bim/bid-deficient mice where only bimappears to impart significantly enhanced survival of DCs (Fig. 3).Thus, it will be intriguing to delineate the precise contributions ofT cell and APC intrinsic apoptotic mechanisms to overall controlof T cells.Specificity of action of BH3-only family members is exquisitely

imparted by differential binding to their partner ligands. Bid can

bind with high affinity to most Bcl-2 family members (except Bcl-

2) (10) but does not appear to be required for shutdown of T cell

responses in acute infections. This is possibly due to the re-

quirement for caspase-8 activation downstream of death receptor

signaling for Bid cleavage. Such a situation arises when a patho-

gen persists in the host long-term and periodically causes immune

reactivation. The increased number of virus-specific T cells in bid/

bim double knockout mice compared with mice lacking Bim alone

suggests that the death receptor apoptotic pathway requires Bid to

eliminate virus-specific T cells under some circumstances. Thus,

Bid appears not to be essential in the Fas-mediated induced T cell

death, but importantly acts to amplify the synergy with Bim and

Fas in regulating of T cell death in the setting of persistent in-

fection.In summary, both the intrinsic (Fas-independent) and extrinsic

(Fas-dependent) pathways regulate the life and death of lympho-

cytes. Our results identify that Bid plays an important regulatory

role in CD8+ T cell death in vivo during persistent infection. The

regulation of T cell death during acute or chronic infections is

ultimately determined by the balance between the proapoptotic

factors and their prosurvival counterparts. The mitochondrial path-

way is regulated by proteins of the Bcl-2 family that dominate

during acute infections, whereas the death receptor pathway

involves membrane receptors of the TNFR1 family, which play an

additional role in chronic infections. In this study, we show that

part of this synergy to eliminate lymphocytes in chronic immune

responses is mediated through Bid to optimize removal of acti-vated CD8+ T cells.

AcknowledgmentsWe thank Drs. T. Kaufmann, P. Jost, P. Bouillet, and V.M. Dixit for gifts of

bid2/2, bim2/2, and bim2/2bid2/2 mice, Dr. F. Battye, V. Lapatis, and C.

Tarlinton for expertise with flow cytometry, and M. Camilleri for expert

animal care and technical assistance.

DisclosuresThe authors have no financial conflicts of interest.

References1. Strasser, A. 2005. The role of BH3-only proteins in the immune system. Nat.

Rev. Immunol. 5: 189–200.2. Green, D. R. 2005. Apoptotic pathways: ten minutes to dead. Cell 121: 671–674.3. Newmeyer, D. D., and S. Ferguson-Miller. 2003. Mitochondria: releasing power

for life and unleashing the machineries of death. Cell 112: 481–490.4. Strasser, A., L. O’Connor, and V. M. Dixit. 2000. Apoptosis signaling. Annu.

Rev. Biochem. 69: 217–245.5. Bouillet, P., and L. A. O’Reilly. 2009. CD95, BIM and T cell homeostasis. Nat.

Rev. Immunol. 9: 514–519.6. Pellegrini, M., G. Belz, P. Bouillet, and A. Strasser. 2003. Shutdown of an acute

T cell immune response to viral infection is mediated by the proapoptotic Bcl-2homology 3-only protein Bim. Proc. Natl. Acad. Sci. USA 100: 14175–14180.

7. Pellegrini, M., P. Bouillet, M. Robati, G. T. Belz, G. M. Davey, and A. Strasser.2004. Loss of Bim increases T cell production and function in interleukin 7receptor-deficient mice. J. Exp. Med. 200: 1189–1195.

8. Fischer, S. F., G. T. Belz, and A. Strasser. 2008. BH3-only protein Puma con-tributes to death of antigen-specific T cells during shutdown of an immune re-sponse to acute viral infection. Proc. Natl. Acad. Sci. USA 105: 3035–3040.

9. Hildeman, D. A., Y. Zhu, T. C. Mitchell, P. Bouillet, A. Strasser, J. Kappler, andP. Marrack. 2002. Activated T cell death in vivo mediated by proapoptotic bcl-2family member bim. Immunity 16: 759–767.

10. Chen, L., S. N. Willis, A. Wei, B. J. Smith, J. I. Fletcher, M. G. Hinds,P. M. Colman, C. L. Day, J. M. Adams, and D. C. Huang. 2005. Differentialtargeting of prosurvival Bcl-2 proteins by their BH3-only ligands allows com-plementary apoptotic function. Mol. Cell 17: 393–403.

11. Kuwana, T., L. Bouchier-Hayes, J. E. Chipuk, C. Bonzon, B. A. Sullivan,D. R. Green, and D. D. Newmeyer. 2005. BH3 domains of BH3-only proteinsdifferentially regulate Bax-mediated mitochondrial membrane permeabilizationboth directly and indirectly. Mol. Cell 17: 525–535.

12. Weant, A. E., R. D. Michalek, I. U. Khan, B. C. Holbrook, M. C. Willingham, andJ. M. Grayson. 2008. Apoptosis regulators Bim and Fas function concurrently tocontrol autoimmunity and CD8+ T cell contraction. Immunity 28: 218–230.

13. Hughes, P. D., G. T. Belz, K. A. Fortner, R. C. Budd, A. Strasser, and P. Bouillet.2008. Apoptosis regulators Fas and Bim cooperate in shutdown of chronic im-mune responses and prevention of autoimmunity. Immunity 28: 197–205.

14. Hutcheson, J., J. C. Scatizzi, A. M. Siddiqui, G. K. Haines, III, T. Wu, Q. Z. Li,L. S. Davis, C. Mohan, and H. Perlman. 2008. Combined deficiency of proa-poptotic regulators Bim and Fas results in the early onset of systemic autoim-munity. Immunity 28: 206–217.

15. Strasser, A., P. J. Jost, and S. Nagata. 2009. The many roles of FAS receptorsignaling in the immune system. Immunity 30: 180–192.

16. Green, D. R. 2008. Fas Bim boom! Immunity 28: 141–143.17. Kaufmann, T., P. J. Jost, M. Pellegrini, H. Puthalakath, R. Gugasyan,

S. Gerondakis, E. Cretney, M. J. Smyth, J. Silke, R. Hakem, et al. 2009. Fatalhepatitis mediated by tumor necrosis factor TNFalpha requires caspase-8 andinvolves the BH3-only proteins Bid and Bim. Immunity 30: 56–66.

18. Li, H., H. Zhu, C. J. Xu, and J. Yuan. 1998. Cleavage of BID by caspase 8mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell 94:491–501.

19. Chen, W., P. A. Calvo, D. Malide, J. Gibbs, U. Schubert, I. Bacik, S. Basta,R. O’Neill, J. Schickli, P. Palese, et al. 2001. A novel influenza A virus mito-chondrial protein that induces cell death. Nat. Med. 7: 1306–1312.

20. Zamarin, D., A. Garcıa-Sastre, X. Xiao, R. Wang, and P. Palese. 2005. Influenzavirus PB1-F2 protein induces cell death through mitochondrial ANT3 andVDAC1. PLoS Pathog. 1: e4.

21. Tschopp, J., M. Irmler, and M. Thome. 1998. Inhibition of fas death signals byFLIPs. Curr. Opin. Immunol. 10: 552–558.

22. Kaufmann, T., L. Tai, P. G. Ekert, D. C. Huang, F. Norris, R. K. Lindemann,R. W. Johnstone, V. M. Dixit, and A. Strasser. 2007. The BH3-only protein bid isdispensable for DNA damage- and replicative stress-induced apoptosis or cell-cycle arrest. Cell 129: 423–433.

23. Bouillet, P., D. Metcalf, D. C. Huang, D. M. Tarlinton, T. W. Kay, F. Kontgen,J. M. Adams, and A. Strasser. 1999. Proapoptotic Bcl-2 relative Bim required forcertain apoptotic responses, leukocyte homeostasis, and to preclude autoimmu-nity. Science 286: 1735–1738.

24. Kupresanin, F., J. Chow, A. Mount, C. M. Smith, P. G. Stevenson, and G. T. Belz.2007. Dendritic cells present lytic antigens and maintain function throughoutpersistent gamma-herpesvirus infection. J. Immunol. 179: 7506–7513.

The Journal of Immunology 7

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

25. Flynn, K. J., G. T. Belz, J. D. Altman, R. Ahmed, D. L. Woodland, andP. C. Doherty. 1998. Virus-specific CD8+ T cells in primary and secondary in-fluenza pneumonia. Immunity 8: 683–691.

26. Belz, G. T., W. Xie, J. D. Altman, and P. C. Doherty. 2000. A previously un-recognized H-2D(b)-restricted peptide prominent in the primary influenza Avirus-specific CD8(+) T-cell response is much less apparent following secondarychallenge. J. Virol. 74: 3486–3493.

27. Belz, G. T., G. M. Behrens, C. M. Smith, J. F. Miller, C. Jones, K. Lejon,C. G. Fathman, S. N. Mueller, K. Shortman, F. R. Carbone, and W. R. Heath.2002. The CD8alpha(+) dendritic cell is responsible for inducing peripheral self-tolerance to tissue-associated antigens. J. Exp. Med. 196: 1099–1104.

28. Henri, S., D. Vremec, A. Kamath, J. Waithman, S. Williams, C. Benoist,K. Burnham, S. Saeland, E. Handman, and K. Shortman. 2001. The dendritic cellpopulations of mouse lymph nodes. J. Immunol. 167: 741–748.

29. McAuley, J. L., F. Hornung, K. L. Boyd, A. M. Smith, R. McKeon, J. Bennink,J. W. Yewdell, and J. A. McCullers. 2007. Expression of the 1918 influenza Avirus PB1-F2 enhances the pathogenesis of viral and secondary bacterialpneumonia. Cell Host Microbe 2: 240–249.

30. McAuley, J. L., K. Zhang, and J. A. McCullers. 2010. The effects of influenza Avirus PB1-F2 protein on polymerase activity are strain specific and do not impactpathogenesis. J. Virol. 84: 558–564.

31. Zamarin, D., M. B. Ortigoza, and P. Palese. 2006. Influenza A virus PB1-F2protein contributes to viral pathogenesis in mice. J. Virol. 80: 7976–7983.

32. Smith, C. M., G. T. Rosa, J. S. May, N. J. Bennett, A. M. Mount, G. T. Belz, andP. G. Stevenson. 2006. CD4+T cells specific for amodel latency-associated antigenfail to control a gammaherpesvirus in vivo. Eur. J. Immunol. 36: 3186–3197.

33. Stevenson, P. G., J. P. Simas, and S. Efstathiou. 2009. Immune control ofmammalian gamma-herpesviruses: lessons from murid herpesvirus-4. J. Gen.Virol. 90: 2317–2330.

34. Belz, G. T., H. Liu, S. Andreansky, P. C. Doherty, and P. G. Stevenson. 2003.Absence of a functional defect in CD8+ T cells during primary murinegammaherpesvirus-68 infection of I-A(b-/-) mice. J. Gen. Virol. 84: 337–341.

35. Sutton, V. R., J. E. Davis, M. Cancilla, R. W. Johnstone, A. A. Ruefli,K. Sedelies, K. A. Browne, and J. A. Trapani. 2000. Initiation of apoptosis bygranzyme B requires direct cleavage of bid, but not direct granzyme B-mediatedcaspase activation. J. Exp. Med. 192: 1403–1414.

36. Waterhouse, N. J., K. A. Sedelies, K. A. Browne, M. E. Wowk, A. Newbold,V.R. Sutton,C. J. Clarke, J.Oliaro, R.K. Lindemann, P. I. Bird, et al. 2005.A centralrole for Bid in granzyme B-induced apoptosis. J. Biol. Chem. 280: 4476–4482.

37. Fischer, K., R. Andreesen, and A. Mackensen. 2002. An improved flow cyto-metric assay for the determination of cytotoxic T lymphocyte activity. J.Immunol. Methods 259: 159–169.

38. Stranges, P. B., J. Watson, C. J. Cooper, C. M. Choisy-Rossi, A. C. Stonebraker,R. A. Beighton, H. Hartig, J. P. Sundberg, S. Servick, G. Kaufmann, et al. 2007.Elimination of antigen-presenting cells and autoreactive T cells by Fas con-tributes to prevention of autoimmunity. Immunity 26: 629–641.

8 Bid AND PERSISTENT INFECTION

by guest on March 27, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from