LFA-1 Regulates CD8� T Cell Activation via T CellReceptor-mediated and LFA-1-mediated Erk1/2Signal Pathways*□S

Received for publication, April 2, 2009, and in revised form, May 5, 2009 Published, JBC Papers in Press, May 29, 2009, DOI 10.1074/jbc.M109.002865

Dan Li, Jeffrey J. Molldrem, and Qing Ma1

From the Section of Transplantation Immunology, Department of Stem Cell Transplantation and Cellular Therapy, University ofTexas M. D. Anderson Cancer Center, Houston, Texas 77030

LFA-1 regulates T cell activation and signal transductionthrough the immunological synapse. T cell receptor (TCR)stimulation rapidly activates LFA-1, which provides uniqueLFA-1-dependent signals to promote T cell activation. How-ever, the detailedmolecular pathways that regulate these pro-cesses and the precise mechanism by which LFA-1 contrib-utes to TCR activation remain unclear. We found LFA-1directly participates in Erk1/2 signaling upon TCR stimula-tion in CD8� T cells. The presence of LFA-1, not ligand bind-ing, is required for the TCR-mediated Erk1/2 signal pathway.LFA-1-deficient T cells have defects in sustained Erk1/2 sig-naling and TCR/CD3 clustering, which subsequently pre-vents MTOC reorientation, cell cycle progression, and mito-sis. LFA-1 regulates the TCR-mediated Erk1/2 signal pathwayin the context of immunological synapse for recruitment andamplification of the Erk1/2 signal. In addition, LFA-1 ligationwith ICAM-1 generates an additional Erk1/2 signal, whichsynergizes with the existing TCR-mediated Erk1/2 signal toenhance T cell activation. Thus, LFA-1 contributes to CD8�

T cell activation through two distinct signal pathways. Wedemonstrated that the function of LFA-1 is to enhance TCRsignaling through the immunological synapse and deliver dis-tinct signals in CD8� T cell activation.

Leukocyte function-associated antigen-1 (LFA-1)2 playsan important role in regulating leukocyte adhesion and T cellactivation (1, 2). LFA-1 consists of the �L (CD11a) and �2(CD18) subunits. The ligands for LFA-1 include intercellularadhesion molecular-1 (ICAM-1), ICAM-2, and ICAM-3 (3).

LFA-1 participates in the formation of the immunologicalsynapse, which regulates T cell activation synergisticallywith TCR engagement. The immunological synapse is a spe-cialized structure that forms between the T cell and the APCor target cell (1, 2, 4). The function of the immunologicalsynapse is to facilitate T cell activation and signal transduc-tion. Mice deficient in LFA-1 (CD11a KO) have defects inleukocyte adhesion, lymphocyte proliferation, and tumorrejection (5–7).Upon TCR stimulation, the nascent immunological syn-

apse is initiated with surface receptor clustering andcytoskeleton rearrangement, then followed by mature syn-apse formation after prolonged stimulation (8, 9). In themature immunological synapse, LFA-1 forms a ring-like pat-tern at the peripheral supramolecular activation cluster(pSMAC), which surrounds the central supramolecular acti-vation cluster (cSMAC) containing TCR/CD3/lipid rafts (10,11). The structure of the mature synapse is stable for hoursand thought to be important for sustained TCR signaling(12–14). LFA-1 functions via pSMAC to stabilize thecSMAC and is associated with the induction of T cell prolif-eration, cytokine production, and lytic granule migrationtoward cSMAC (1, 15). Although LFA-1-containing pSMACis self-evident in lipid bilayer systems and cell lines, whetherit is required for T cell activation under physiological condi-tions remains controversial (15).TCR stimulation rapidly induces the functional activity of

LFA-1, which then provides unique LFA-1-dependent sig-nals to promote T cell activation (16). The process can bedivided into two steps. First, the intracellular signaling fromTCR regulating LFA-1 activation is known as “inside-out”signaling; second, activated LFA-1, as a signaling receptor,can feedback to transduce the intracellular signal, the “out-side-in” signaling (1, 17). It is widely accepted that TCR stim-ulation activates LFA-1 through affinity and/or avidity reg-ulation, as supported by increased adhesion to ICAM-1 andpSMAC formation (16, 17). The “inside-out” signal processhas been investigated extensively (18–21). The TCR proxi-mal signal molecules, Lck, ZAP-70, and PI3K, are known tobe important for TCR signaling to LFA-1 activation (22–26).The molecular mechanisms of LFA-1 “outside-in” signalinghave been explored only recently. Perez et al. (27) have dem-onstrated that LFA-1 and ICAM-1 ligation activates thedownstream Erk1/2 MAPK signaling pathway upon TCRstimulation, which ultimately leads to the qualitative modu-

* This work was supported by American Cancer Society Grant RSG-08-183-01-LIB (to Q. M.).

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Fig. S1.

1 To whom correspondence should be addressed: Section of TransplantationImmunology, Dept. of Stem Cell Transplantation and Cellular Therapy, Uni-versity of Texas M.D. Anderson Cancer Center, Unit 900, 1515 HolcombeBlvd., Houston, TX 77030. Tel.: 713-563-3327; Fax: 713-563-3364; E-mail:qma@mdanderson.org.

2 The abbreviations used are: LFA, leukocyte function-associated antigen;TCR, T cell receptor; Ab, antibody; FACS, fluorescent-activated cell sort-ing; IL, interleukin; IFN, interferon; WT, wild type; KO, knockout; Erk,extracellular signal-regulated kinase; MTOC, microtubule-organizingcenter; ICAM, intercellular adhesion molecular; pSMAC, peripheralsupramolecular activation cluster; cSMAC, central supramolecular acti-vation cluster; MAPK, mitogen-activated protein kinase; MFI, mean flo-rescence intensity; PI, propidium iodide; CTB, cholera toxin subunit B;TNF, tumor necrosis factor.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 284, NO. 31, pp. 21001–21010, July 31, 2009© 2009 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

JULY 31, 2009 • VOLUME 284 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21001

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

lation of CD4� T cell activation through distinct LFA-1-de-pendent signals. Another recent study provided compellingevidence that LFA-1 reshapes the RasMAPK pathway down-stream of TCR (28). However, the detailed molecular path-ways that regulate these processes are poorly defined. Espe-cially, the evidence in support of a distinctive role for LFA-1in the T cell signaling pathway has lagged behind; whetherthe function of LFA-1 is to enhance TCR signaling throughthe immunological synapse and/or deliver distinct signal inT cell activation and whether LFA-1 is indispensable for ormerely assists the existing TCR signal pathway. Further-more, whether and how TCR proximal signal molecules reg-ulate LFA-1 function remains unknown. Further studies arerequired to understand the LFA-1 and TCR signalingnetwork.In this study, we found that LFA-1 directly participates in

CD8� T cell activation. Upon TCR stimulation, LFA-1 reg-ulates both TCR-mediated and LFA-1-mediated Erk1/2 sig-nal pathways. First, the presence of LFA-1, not ligand bind-ing, is required for the sustained Erk1/2 signaling and TCR/CD3 clustering on the surface of CD8� T cells, subsequentlyleading to MTOC reorientation, cell cycle progression, andmitosis. Second, LFA-1 ligation with ICAM-1 enhancesErk1/2 signaling, which promotes T cell activation withincreased IL-2 production and cell proliferation. This LFA-1-mediated Erk1/2 signal pathway integrates with the exist-ing TCR-mediated Erk1/2 signal pathway to enhance T cellactivation.

EXPERIMENTAL PROCEDURES

Animals—C57BL/6 mice were purchased from the AnimalProduction Area at NCI Frederick. CD11a KO mice (C57BL/6background) were kindly provided by Dr. Christie Ballantyne(Baylor College of Medicine). The animal experiments wereapproved by the Institutional Animal Care and Use Committeeat University of Texas M. D. Anderson Cancer Center.Reagents—The monoclonal antibodies specific to mouse

CD3 (145-2C11), CD4 (H129.19), CD8 (53-6.7), CD28 (37.51),IL-2 (JES6-5H4), IFN-� (XMG1.2), and Ki-67 (B56) were fromBDBiosciences (SanDiego, CA); TNF-� (MP6-XT22)was fromeBioscience (San Diego, CA), and TCR-� was from InvitrogenDetection Technologies (Carlsbad, CA). The recombinantmouse ICAM-1/Fc chimera was fromR&D Systems (Minneap-olis, MN). The rabbit anti-mouse phospho-p44/22 MAPK(Thr-202/Tyr-204) antibody was obtained from Cell SignalingTechnologies. The secondary Abs AlexaFluor 594 goat anti-hamster IgG (H�L), AlexaFluor 647 goat anti-rabbit IgG, Alex-aFluor 647 goat anti-rat IgG, AlexaFluor 647 cholera toxin sub-unit B (CTB), and Prolong Anti-FadeTM mounting mediumwere purchased from Invitrogen Detection Technologies. Theanti-�-tubulin (TUB2.1) was purchased from Sigma-Aldrich.The piceatannol, LY294002, cytochalasin D, and latrunculinwere from EMD Biosciences (Darmstadt, Germany), and PP-1was from Biomol (Plymouth Meeting, PA).Cell Isolation and Stimulation—The single cell suspension

was prepared from spleen and lymphnodes fromC57BL/6miceor CD11a KO mice by a standard method. CD8� T cells werepurified with mouse CD8a� T Lymphocyte Enrichment

Set-DM from BD Biosciences (San Diego, CA). In brief, 5 �l ofbiotin-antibody mixture including biotin-conjugated mono-clonal antibodies against CD4 (GK1.5), CD11b (M1/70),CD45R/B220 (RA3–6B2), CD49b (HM�2), and TER-119/erythroid cells (TER-119) were mixed with 1 � 106 cells for 10min on ice. Then, 5�l of the BDTMIMag Streptavidin ParticlesPlus-DMwere added to the single cell suspension, and CD8� Tcells were negatively selected with the BDTM IMagnet. Thepurified CD8� T cells were diluted to 1 � 106 cells/ml in RPMI1640 medium in 24-well microtiter plates and stimulated withcoated anti-CD3 antibody in the presence or absence of coatedICAM-1 for indicated times.Intracellular Staining—The intracellular staining was per-

formed with the Fix & Perm cell permeabilization reagentsaccording to the manufacturer’s protocol (Invitrogen) with aminormodification. For each sample, 1� 106 cellswerewashedand resuspended in FACS buffer (phosphate-buffered saline,0.1%NaN3, 1% fetal bovine serum). Cells were first incubatedwith various cell surface markers for 20 min, then washed andresuspended in reagent A (fixation medium). The rabbit anti-mouse phospho-p44/22MAPK antibodywas diluted in reagentB (permeabilization medium) and added onto cells. Then, cellswere washed and incubated with secondary antibody Alex-aFluor 647 goat anti-rabbit IgG. For intracellular IL-2, IFN-�,and TNF-� staining, 10 �g/ml Brefeldin A was added for thelast 8 h of stimulation. To detect intracellular cytokine produc-tion, cells were fixed (reagent A) and permeabilized (reagent B),then stained with AlexaFluor 647-conjugated IL-2, PE-Cy7-conjugated IFN-�, and Pacific Blue-conjugated TNF-� anti-body.Multicolor flow cytometry data were acquired with FAC-SCalibur (BD Biosciences) and analyzed with FlowJo software(Tree Star, Ashland, OR).Fluorescence Microscopy and Image Analysis—Fluorescence

staining and image analysis of CD8� T cells were prepared asdescribed below. For cytoskeleton inhibition experiments, cellswere pretreated with 100 M cytochalasin D or 100 �M latruncu-lin A. For lipid raft experiments, cells were pretreated with 25�g/mlAlexaFluor 647-conjugatedCTB.After stimulation, cellswere fixed with 4% paraformaldehyde, then stained with spe-cific monoclonal antibodies conjugated with fluorescence dyes.The secondary antibody goat anti-hamster IgG AlexaFluor 594(1:400) was used to observe the localization of CD3 on the cellsurface. The anti-mouse CD8 AlexaFluor 488 (1:100), anti-mouse TCR-� AlexaFluor 647 (1:50), and anti-mouse CD11aAlexaFluor 488 (1:400) were used for cell surface staining. Therabbit anti-mouse phospho-p44/22MAPK followed by second-ary antibody AlexaFluor 647 goat anti-rabbit IgG in permeabi-lization medium were used to detect p-Erk1/2. The anti-�-tu-bulin Cy3 (1:200) was used to stainMTOC. After washing threetimes in phosphate-buffered saline, the cells were loaded ontopoly-L-lysine-coated slides andmounted ProlongAnti-FadeTM.Fluorescence images were acquired using a Leica TCS SE RSspectral laser-scanning confocal microscope (Leica Microsys-tems,Wetzlar, Germany) with the appropriate filters and a�63objective lens. Image analysis was performed using Leica Con-focal Software (LeicaMicrosystems). Cells were examined withz stacks of 0.2 �m and at least 20 iterations.

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

21002 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 31 • JULY 31, 2009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

RESULTS

LFA-1 Participates in Both TCR-mediated and LFA-1-medi-ated Erk1/2 Signaling—TCR stimulation and LFA-1 ligationcan induce the activation of the Erk1/2 MAPK signal pathwayin CD4� T cells and Jurkat cells (27, 29). Using the phospho-Erk1/2 (p-Erk1/2) flow cytometry assay and CD11a KO mice,we examined the role of LFA-1 andTCR stimulation inCD8�Tcell activation. As shown in Fig. 1A (upper panel), purifiedCD8� T cells were stimulated with coated anti-CD3 antibody,the p-Erk1/2 activity was detected in 40.1% WT and 37.1%CD11a KO CD8� T cells after 5 min of stimulation. The mean

florescence intensity (MFI) was 71.3 and 50.7, respectively. Thep-Erk1/2 activity in WT (left panel) and CD11a KO (rightpanel) cells peaked at 5 min, and the signal subsided to back-ground levels within 60min. The percentage of p-Erk1/2� cellswas similar in WT and CD11a KO CD8� T cells with CD3antibody stimulation. As shown in Fig. 1A (lower panel),ICAM-1 alone did not induce p-Erk1/2, and there was noapparent difference in p-Erk1/2 activity in the presence orabsence of ICAM-1.It has been demonstrated previously that continuous TCR

stimulation is required for signal amplification during T cell

FIGURE 1. LFA-1 ligation and TCR stimulation induce Erk1/2 signaling in CD8� T cells. Purified mouse CD8� T cells were stimulated with various reagentsfor the indicated times. Cells were harvested and stained with CD8 antibody and then fixed and permeabilized before incubation with the phospho-p44/22MAPK antibody. The data were collected by flow cytometry and analyzed with FlowJo software. A, phospho-Erk1/2 activity in WT (right panel) and CD11a KO(left panel) CD8� T cells. The p-Erk1/2 was measured after 0, 5, 15, 30, and 60 min of stimulation. The upper panel shows representative dot plots after 5 min ofstimulation, and the lower panel shows the percentage of p-Erk1/2-positive cells at each time point. Results are plotted as the mean of three independentexperiments with S.D. The percentage and MFI of each population are displayed. B, phospho-Erk1/2 activity in CD8� T cells stimulated for 12 h. The panels fromleft to right are: WT cells (top) with various stimulation (unstimulated control, CD3 antibody, CD3 antibody and ICAM-1, ICAM-1) and CD11a KO cells (bottom)with various stimulation (unstimulated control, CD3 antibody, CD3 antibody and ICAM-1). The numbers in the quadrants represent the percentage of eachpopulation. The MFI of cell populations with relatively high p-Erk1/2 (p-Erk1/2High) and low p-Erk1/2 (p-Erk1/2Low) activities are displayed in circles. The data arerepresentative of at least three independent experiments.

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

JULY 31, 2009 • VOLUME 284 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21003

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

21004 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 31 • JULY 31, 2009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

activation (2, 4, 8–9). To further investigate whether LFA-1plays a role in sustained Erk1/2 signaling in CD8� T cells, wemeasured p-Erk1/2 activity after prolonged TCR cross-linkingwith CD3 antibody for 6–18 h. As shown in Fig. 1B, there weretwo distinct subsets of cells with intracellular p-Erk1/2 activityin the presence of ICAM-1: one with relatively high amounts ofp-Erk1/2(p-Erk1/2High) and one with relatively low amounts ofp-Erk1/2(p-Erk1/2Low). The p-Erk1/2 activity measured bythe MFI in the p-Erk1/2High subset was 17 times that of thep-Erk1/2Low subset (815 versus 50). However, the CD3 anti-body stimulation alone only induced the p-Erk1/2Low subset.The p-Erk1/2 activity of cells stimulated with ICAM-1 in theabsence of TCR cross-linking was similar to that of theunstimulated control without any detectable p-Erk1/2 activity,which indicated that signal from LFA-1 ligation with ICAM-1alone could not activate the Erk1/2 signal pathway. In addition,only the p-Erk1/2High but not the p-Erk1/2Low could be inhib-ited with an LFA-1-blocking antibody M17/4, which abolishedLFA-1 binding to ICAM-1 (data not shown). Thus, LFA-1 liga-tion with ICAM-1 provided a unique LFA-1-dependent Erk1/2signal upon TCR stimulation. Furthermore, there was only aminimal amount of detectable p-Erk1/2 activity in CD11a KOcells upon continuous TCR stimulation for 12 h (Fig. 1B),although a significant amount of p-Erk1/2 was detected 5 minafter CD3 antibody stimulation (Fig. 1A). Therefore, thep-Erk1/2Low subset is generated with continuous TCR stimula-tion in the presence of LFA-1.In summary,we found that LFA-1 contributes toCD8�Tcell

activation in two ways: first, the presence of LFA-1 is requiredfor the p-Erk1/2 activity induced by TCR stimulation, which isrepresented by the p-Erk1/2Low subset (TCR-mediated Erk1/2signaling); second, the ligation of LFA-1 with ICAM-1 signifi-cantly enhances intracellular Erk1/2 signaling, which is repre-sented by the p-Erk1/2High subset (LFA-1-mediated Erk1/2 sig-naling). Although TCR stimulation and LFA-1 ligation canactivate Erk1/2 signaling inCD8�T cells, these signal pathwaysdepend on the presence of LFA-1 on the cell surface.The Presence of LFA-1 Is Essential for TCR-mediated Erk1/2

Signaling and TCR/CD3Clustering—To further investigate themechanism of LFA-1 in regulating TCR-induced Erk1/2 activ-ity, we used fluorescence microscopy to examine the subcellu-lar location and dynamics of p-Erk1/2 activity. As shown in Fig.2A (upper panel), p-Erk1/2 was enriched proximal to themem-brane and associated with the prominent CD3macro-cluster in

the WT CD8� T cells after 30 min of TCR stimulation. Therelative location and quantity of p-Erk1/2 (blue line) and CD3(red line) were plotted on the right side of the overlay image,with two peaks of p-Erk1/2 and CD3 co-localized to the cellmembrane. In the CD11a KO cells, the predominant patternwas different with multiple CD3/p-Erk1/2 micro-clusters dis-tributed on the cell membrane. However, the p-Erk1/2 activitywas evident and co-localized with CD3 proximal to the cellmembrane. This is consistent with the data in Fig. 1A showingthat p-Erk1/2 activity in WT and CD11a KO cells was compa-rable. After 8–12 h of stimulationwith CD3 antibody, p-Erk1/2remained co-localized with CD3 macro-clusters in the WTCD8� T cells (Fig. 2A, middle panel). Furthermore, p-Erk1/2was translocated inside of the cell, and a significant amount ofp-Erk1/2 activity was detected in the cytosol, which indicatedthe continuous recruitment and amplification of p-Erk1/2upon TCR stimulation. However, the p-Erk1/2 activity wasbarely detectable proximal to the membrane in CD11a KOcells, althoughCD3micro-clusters remained evenly distributedacross the cell surface. In addition, the p-Erk1/2 signal inside ofthe cells was minimal and at background levels. Although thep-Erk1/2 activity can be transiently initiated upon TCR stimu-lation in CD11a KO cells, the sustained Erk1/2 signalingrequires the presence of LFA-1.The cytoskeleton remodeling and lipid rafts have been dem-

onstrated to be involved in TCR clustering and immunologicalsynapse formation (1, 30, 31). To examine the role of LFA-1 inregulating CD3 macro-cluster formation, CD8� T cells werepreincubated with CTB before CD3 stimulation. As shown inFig. 2A (bottom panel), a CD3 macro-cluster was associatedwith the lipid microdomain enriched with GM1 gangliosides inWT cells upon TCR stimulation for 30 min. The CD3 macro-clusterswere detected in about 50%of theWTcells after 30minof TCR stimulation (left panel in Fig. 2B). In the CD11a KOcells, there was a significant reduction of cells containing CD3macro-clusters (18.75%), which was similar to WT control(13.77%), WT cells pretreated with actin polymerization inhib-itor cytochalasin D (11.64%), or latrunculin A (18.61%). After12 h of TCR stimulation (right panel in Fig. 2B), CD3 macro-clusters presented on 80.72% of the wild-type cells comparedwith 10.69% CD11a KO cells. Thus, LFA-1 plays an essentialrole in CD3macro-cluster formation, whichmaintains the sus-tained Erk1/2 signaling. The presence of LFA-1 on the cell sur-

FIGURE 2. The presence of LFA-1 is required for sustained Erk1/2 signaling and CD3 macro-cluster formation induced by TCR stimulation. CD8� T cellsfrom WT or CD11a KO mice were stimulated with coated anti-CD3 antibody for the indicated times. A, LFA-1 is required for sustained Erk1/2 signaling inducedby TCR stimulation. Cells were harvested and stained with CD8 (green), CD3 (red), and phospho-p44/22 MAPK antibodies (blue). For lipid raft experiments, cellswere pretreated with CTB (blue) before the stimulation, and then stained with CD3 (red) and CD8 antibodies (green). Images were acquired using Leica TCS SERS spectral laser-scanning confocal microscopy with the appropriate filters and a �63 oil objective lens. Images in the middle sections of the cells werecaptured, and each set of images was representative of at least 50 cells examined in three independent experiments. In the merged images, the arrowheadsindicate the enrichment and colocalization of CD3 and p-Erk1/2 on the cell membrane. In the fluorescence intensity profiles of p-Erk1/2 (blue line) and CD3 (redline), cells were quantified using the Quantify Package of the Leica Confocal Software (version 2.61 of LCS Lite), with sampling at every 1.16E-08 �m distanceand averaged for the mean total fluorescence intensity in arbitrary fluorescence units. The vertical lines represent the boundaries of the cell membrane andintracellular fluorescence signals. B, LFA-1 is required for CD3 macro-cluster formation induced by TCR stimulation. CD8� T cells from WT and CD11a KO micewere quantified after stimulation for 30 min (left panel) or 12 h (right panel). In addition, WT cells were stimulated for 30 min in the presence or absence ofcytochalasin D (100 �M) or latrunculin A (100 �M). AlexaFluor 594-conjugated secondary antibody to CD3 and AlexaFluor 488-conjugated CD8 antibody wereused to detect CD3 macro-clustering (capping) on CD8� T cells under a confocal microscope. The plot represents the percentages of cells with CD3 cappingwith three independent experiments analyzed. The samples are the WT unstimulated (n � 208), WT with CD3 antibody in the presence of cytochalasin D (n �567) or latrunculin A (n � 274), WT with CD3 antibody (n � 875), CD11a KO with CD3 antibody (n � 271), respectively. A significance level of p � 0.001 wasdetermined using the WT with CD3 antibody stimulation sample as the reference.

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

JULY 31, 2009 • VOLUME 284 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21005

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

21006 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 31 • JULY 31, 2009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

face is required for TCR-mediated Erk1/2 activity and TCR/CD3 clustering.Requirement of LFA-1 for MTOC Reorientation, Cell Cycle

Progression, and Mitosis Induced by TCR Stimulation—UponTCR stimulation, MTOC is reoriented to the immunologicalsynapse through cytoskeleton reorganization, and it has beenproposed that microtubules anchor in the pSMAC (32–34). Inaddition, Erk1/2 signaling is required for the polarization ofMTOC in NK cells (35). Because the presence of LFA-1 isrequired for the Erk1/2 signaling in CD8� T cells, we examinedMTOC localization in WT and CD11a KO cells. Upon TCRstimulation for 30 min, there were slightly increased MTOCsreoriented to the proximal cell membrane in WT cells com-paredwithCD11aKO (data not shown). After TCR stimulationfor 12 h, theMTOCwas reoriented to theCD3macro-cluster inWT cells, whereas the MTOC remained inside of the CD11aKO cells (upper panel in Fig. 3A). As shown in Fig. 3B, therewere 77.7% WT cells with polarized MTOC in comparison toonly 17.24% CD11a KO cells. After 48 h of stimulation, theWTCD8� T cells were mostly premitotic blasts with a single polar-ized MTOC (74.78%) and about 10.81% mitotic cells with twopolarizedMTOCs (lower panel in Fig. 3,A andC).We observedthe dissolution of cytoskeleton structure including MTOC andsignificant apoptosis in CD11a KO cells after 48 h of stimula-tion (data not shown). Thus, in the absence of LFA-1, theMTOC polarization induced by TCR stimulation is defective.We further investigated whether the absence of LFA-1 and

Erk1/2 signaling plays a role in cell cycle entry andmitosis. Thepercentage of viable cells in the G0, G1, S, and G2/M phaseswere determined by double staining of Ki-67 and PI (36, 37). Asshown in Fig. 3D, there were 5.02, 3.03, and 2.63% wild-typecells in G1, S, and G2/M phases respectively in WT cells after48 h of stimulation. Although CD11a KO cells could progressthrough the cell cycle, there were reduced cells in G1, S, andG2/M phases (3.33, 1.05, and 1.08%). In addition, there wereincreased dying cells in CD11a KO cells in comparison to WT(16.7% versus 42.1%). Collectively, CD11a KOCD8� T cells canenter the cell cycle upon TCR stimulation but at a much slowerrate than WT.LFA-1-mediated Erk1/2 Signaling Enhances the Activation

and Proliferation of CD8�TCells—We further investigated thefunctional consequences of LFA-1 ligation with ICAM-1 inCD8� T cell activation. Intracellular IL-2, TNF-�, and IFN-�weremeasured simultaneously 12 h after stimulation with CD3antibody or CD3 plus ICAM-1. As shown in Fig. 4A (upperpanel), there were two subsets of activated CD8� T cells withrelatively high and low levels of intracellular IL-2 in the pres-ence of ICAM-1. The total amount of intracellular IL-2 was

significantly increased in the presence of ICAM-1 comparedwith CD3 antibody stimulation alone (MFI: 133 versus 18), asshown in the lower panel of Fig. 4A. Thus, the activation of theErk1/2 signal pathway by LFA-1 ligation enhances IL-2 produc-tion in CD8� T cells. In addition, the cells positive for intracel-lular IL-2 also coexpressed TNF-� and IFN-�. Whereas theamount of TNF-� remained the same in the presence ofICAM-1, the production of IFN-�was significantly increased asshown in Fig. 4A (lower panel).Subsequently, we examined the effect of LFA-1 ligation on

CD8� T cell proliferation. As shown in Fig. 4B (upper panel),there were similar amounts of cells enteringG1 phase after 72 hof stimulation with CD3 antibody in the absence or presence ofICAM-1 (13.6 versus 13.9). However, LFA-1 ligation withICAM-1 produced a higher percentage of cells in the S phasethan did CD3 antibody stimulation alone (7.4 versus 6.43), anda similar effect was seen in cells entering G2-M phase (3.6 ver-sus 2.22). Overall, there were significantly increased cells incycle (S�G2-M) in the presence of ICAM-1 (Fig. 4B, lowerpanel). Thus, LFA-1 ligation with ICAM-1 promotes cell cycleentry of CD8� T cells.

DISCUSSION

Wehave investigated themolecularmechanisms of LFA-1 inCD8�Tcell activation.We found that LFA-1 regulatesCD8�Tcell activation through two distinct signal pathways. UponCD3stimulation, the presence of LFA-1 on the cell surface isrequired for TCR-mediated Erk1/2 signaling. In addition,LFA-1 ligation with ICAM-1 produces a unique LFA-1-medi-ated Erk1/2 signal, which integrates with the TCR-mediatedErk1/2 signal for optimal T cell activation. Our data demon-strated the differential contribution of TCR stimulation andLFA-1 ligation in the activation of the Erk1/2 pathway.We have shown that LFA-1 plays an essential role in the

sustained Erk1/2 signaling induced by TCR stimulation. LFA-1is not required to initiate theTCR-mediatedErk1/2 signal path-way because Erk1/2 signaling occurs in the absence of LFA-1.However, the presence of LFA-1 is required for TCR/CD3mac-ro-cluster formation and sustained Erk1/2 signaling. Subse-quently, the CD8� T cells commit to the activation programsuch as cytokine production and cell cycle progression. Wedemonstrated that the TCR/CD3macro-cluster is co-localizedwith p-Erk1/2 activity. Newly formedTCR/CD3micro-clustersare actively transported and reorganized into themacro-clusterupon TCR stimulation (1, 4, 38). The absence of LFA-1 resultsin defective TCR/CD3 macro-cluster formation, which has astructural and functional similarity to the cSMAC. The matureimmunological synapse is a long-lived structure, and the

FIGURE 3. Requirement of LFA-1 for MTOC polarization, cell cycle entry, and mitosis induced by TCR stimulation. A–C, CD8� T cells from WT or CD11a KOmice were collected and stimulated with coated anti-CD3 antibody for 12 h and 48 h. CD8 (blue), CD3 (red), and �-tubulin antibodies (green) were used to stainthe cells. Images were acquired using confocal microscopy with the appropriate filters, and cells were scanned with z stacks of 0.2 �m for the presence of MTOC.Representative images of premitotic cells with single MTOC proximal to cell membrane (upper panel) and mitotic cells with two opposite MTOCs (lower panel)are shown (A). The percentages of WT and CD11a KO cells with single polarized MTOC after 12 h of stimulation with CD3 antibody were quantified; asignificance level of p � 0.001 was determined using the unstimulated WT sample as the reference (B). The percentages of wild-type cells with single polarizedMTOC and two opposite polarized MTOCs after 48 h of stimulation with CD3 antibody were quantified, respectively; a significance level of p � 0.001 wasdetermined using the single MTOC sample as the reference (C). D, cell cycle analyses. CD8� T cells were fixed and permeabilized after 24 h of stimulation withcoated anti-CD3 antibody. Cells were stained with CD8 and Ki-67 antibodies. Cell viability was monitored by PI staining. Cells in the G0, G1, S, and G2/M phasesof the cell cycle were determined by double staining for expression of Ki-67 and DNA content as indicated. The percentages of cells in G1 and S-G2/M werequantified. Results shown are representative data and plotted as the mean of three independent experiments with S.D.

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

JULY 31, 2009 • VOLUME 284 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21007

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

21008 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 31 • JULY 31, 2009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

cSMAC is required for the maintenance of sustained signaling(1–2, 4). In the absence of LFA-1 and the TCR/CD3 macro-cluster as an anchor, the local concentration of p-Erk1/2 is notmaintained and amplified over time, and the MTOC does notpolarize toward the immunological synapse. This results inincomplete CD8� T cell activation.

The presence of LFA-1 is essential for sustained TCR-medi-ated Erk1/2 signaling (p-Erk1/2Low), and LFA-1-mediatedErk1/2 signaling (p-Erk1/2High) promotes the optimal activa-tion of CD8� T cells. Both signal pathways depend on TCR andLFA-1. To investigate whether the LFA-1-mediated Erk1/2 sig-nal pathway is directly regulated by the proximal signals fromTCR stimulation, we used the pharmacological inhibitors Lck(PP1), PI3K (LY294002), andZAP70 (piceatannol). As shown insupplemental Fig. S1A, preincubation of CD8� T cells withPP1, piceatannol, or LY294002 blocked TCR-mediated Erk1/2activity (p-Erk1/2Low), regardless of the absence (upper panel)or presence (lower panel) of ICAM-1. However, the p-Erk1/2High subset resulting from LFA-1 ligation with ICAM-1 wasnot affected by LY294002 and piceatannol, but partiallyimpaired by PP1with 66% reduction of cells with p-Erk1/2Highactivity (supplemental Fig. S1B). These data indicated the exist-ence of two signal pathways: the TCR-mediated Erk1/2 signalpathway (p-Erk1/2Low) depends on Lck, ZAP-70, and PI3K,whereas the LFA-1-mediated Erk1/2 signal pathway (p-Erk1/2High) is dependent on Lck but independent of ZAP-70 andPI3K. Thus, TCR-mediated and LFA-1-mediated Erk1/2 signalpathways are differentially regulated by proximal signals fromTCR activation.Lck, ZAP-70, and PI3K function sequentially to initiate TCR

signal transduction and activate the downstreamErk1/2MAPKpathway (15, 23). In addition, the TCR-dependent Lck/Zap70/PI3K pathway can trigger actin cytoskeleton rearrangementand recruit LFA-1 to the peripheral zone of the pSMAC (17,21).We have demonstrated that the TCR-mediated Erk1/2 sig-nal pathway depends on Lck, ZAP-70, and PI3K. Furthermore,cytochalasin D and latrunculin A, inhibitors that disrupt thecytoskeleton, prevent sustained Erk1/2 activity and TCR/CD3clustering on the cell surface. We propose that the function ofLFA-1 in the TCR-mediated Erk1/2 signal pathway is as fol-lows: TCR signaling recruits and activates LFA-1 via cytoskel-eton rearrangement; subsequently, LFA-1 functions viapSMAC to stabilize the cSMAC and sustain Erk1/2 signalingin T cell activation. Thus, LFA-1 regulates the TCR-medi-ated Erk1/2 signal pathway in the context of the immunolog-ical synapse.The LFA-1-mediated Erk1/2 signal pathway, which is

Zap70- and PI3K-independent but partially Lck-dependent,requires TCR stimulation and LFA-1 ligation for optimal

CD8� T cell activation. The p-Erk1/2High population indi-cates that LFA-1 ligation with ICAM-1 can significantlyenhance intracellular Erk1/2 MAPK signaling in the pres-ence of TCR stimulation. This is consistent with the previousobservation that LFA-1 signaling cooperates with TCR stim-ulation to augment phosphorylation of Erk1/2 and lower theactivation threshold in CD4� T cells (27). Here, we demon-strated that LFA-1 influences CD8� T cell activation by pro-viding a distinct signal through ligation with ICAM-1, whichintegrates with the TCR-mediated Erk1/2 signal to enhanceT cell activation.LFA-1 is constitutively expressed on the surface of leuko-

cytes in an inactive state (39–41). It is controversial whetherthe affinity or avidity state of LFA-1 is involved in the inside-out and outside-in signals. A recent publication demon-strated that cytoskeletal regulation couples LFA-1 confor-mational changes to receptor lateral mobility and clustering;thus, both affinity and avidity states of LFA-1 provide regu-lation of receptor function on lymphocytes (42). Previousstudies reported that LFA-1 ligation can co-stimulate T cellsfor IL-2 production, cell proliferation, and Th1 polarization(27, 43–44). However, the detailed molecular pathways thatregulate these processes are not defined. In particular, evi-dence in support of a distinctive role for LFA-1 activation inT cell signaling pathway is lacking. Based on our data, aplausible model is derived as follows: the TCR signaling toLFA-1 (inside-out) activates both the avidity and affinity ofLFA-1; the avidity regulation of LFA-1 in the absence ofICAM-1 ligand is mediated through pSMAC formation,which leads to the maintenance of the stable immunologicalsynapse for the TCR-mediated Erk1/2 signal pathway. This isconsistent with a previous report that TCR stimulation acti-vates Rap1, which leads LFA-1 to rapidly accumulate atimmunological synapses through spatial redistribution (45).Furthermore, integrin-activating agonists have been shownto cause LFA-1 to move into lipid rafts (30). In addition, thepolarization and clustering of LFA-1 can be observed when Tcells are stimulated with CD3 antibody (2. 4). The TCR sig-naling to LFA-1 also changes the affinity state of LFA-1simultaneously. Subsequently, the ligation of high affinityLFA-1 with ICAM-1 transduces the outside-in signal. Theligation between activated LFA-1 and ICAM-1 providesextra strength to stabilize the immunological synapse andaccessory signals to accelerate T cell activation. Thus, ourresults suggest that TCR controls the avidity and affinity ofLFA-1, which are coupled to provide regulation of CD8� Tcell activation.In summary, we found that LFA-1 contributes to CD8� T

cell activation through two distinct Erk1/2 signal pathways.

FIGURE 4. LFA-1 ligation with ICAM-1 promotes T cell activation and proliferation. A, intracellular cytokine production in CD8� T cells with LFA-1 ligation.CD8� T cells were stimulated with coated anit-CD3 antibody with or without coated ICAM-1 for 12 h. Intracellular IL-2, TNF-�, and IFN-� were measuredsimultaneously. The upper panel shows the representative dot plots of intracellular IL-2 production. The IL-2� cells are displayed in circles. The lower panelshows the MFI of each cytokine quantified. Results are plotted as the mean of three independent experiments with S.D.; a significance level of p � 0.05 wasdetermined using CD3 antibody stimulation sample as the reference. B, cell cycle analysis of CD8� T cells with LFA-1 ligation. CD8� T cells were stimulated withcoated anti-CD3 antibody with or without coated ICAM-1 for 72 h. The percentage of viable cells in the G0, G1, S, and G2/M phases of the cell cycle weredetermined by double staining of Ki-67 and PI. Values of PI lower than 500 were considered as cells in G0/G1 (1n DNA); values from 500 –700 were consideredas cells in S (1n-2n DNA); values higher than 700 were considered as cells in G2/M (2n DNA). Low Ki-67 signals were considered as cells in G0; Ki-67 signals with2n DNA were considered as cells in G1. The percentages of cells in G1 and S�G2/M were quantified. Results are representative data and plotted as the meanof three independent experiments with S.D.; a significance level of p � 0.05 was determined using CD3 antibody stimulation sample as the reference.

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

JULY 31, 2009 • VOLUME 284 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

The first Erk1/2 signal is delivered through the TCR byrecruiting LFA-1 to the immunological synapse. This TCR-mediated Erk1/2 signal pathway is required for CD8� T cellactivation. The second Erk1/2 signal is delivered throughLFA-1/ICAM-1 binding. This LFA-1-mediated Erk1/2 signalpathway induces optimal activation of CD8� T cells.

Acknowledgments—We thank Dr. Christie Ballantyne for providingthe CD11a-deficient mice and Dr. Tomasz Zal for technical assist-ance and discussion on microscopy. The animal experiments wereapproved by the Institutional Animal Care and Use Committee atUniversity of Texas M. D. Anderson Cancer Center. We have no con-flicting financial interests.

REFERENCES1. Dustin, M. L., Bivona, T. G., and Philips, M. R. (2004) Nat. Immunol. 5,

363–3722. Huppa, J. B., and Davis, M. M. (2003) Nat. Rev. Immunol. 3, 973–9833. Springer, T. A. (1994) Cell 76, 301–3144. Lin, J., Miller, M. J., and Shaw, A. S. (2005) J. Cell Biol. 170, 177–1825. Ding, Z.M., Babensee, J. E., Simon, S. I., Lu, H., Perrard, J. L., Bullard, D. C.,

Dai, X. Y., Bromley, S. K., Dustin, M. L., Entman, M. L., Smith, C. W., andBallantyne, C. M. (1999) J. Immunol. 163, 5029–5038

6. Schmits, R., Kundig, T. M., Baker, D. M., Shumaker, G., Simard, J. J.,Duncan, G., Wakeham, A., Shahinian, A., van der Heiden, A., Bachmann,M. F., Ohashi, P. S., Mak, T. W., and Hickstein, D. D. (1996) J. Exp. Med.183, 1415–1426

7. Berlin-Rufenach, C., Otto, F., Mathies, M., Westermann, J., Owen, M. J.,Hamann, A., and Hogg, N. (1999) J. Exp. Med. 189, 1467–1478

8. Dustin, M. L. (2002) Arthritis Res. 4, S119–S1259. Dustin, M. L. (2005) Semin. Immunol. 17, 400–41010. Monks, C. R., Freiberg, B. A., Kupfer H., Sciaky, N., and Kupfer, A. (1998)

Nature 395, 82–8611. Grakoui, A., Bromley, S. K., Sumen, C., Davis, M. M., Shaw, A. S., Allen,

P. M., and Dustin, M. L. (1999) Science 285, 221–22712. Dustin, M. L. (2003) Ann. N.Y. Acad. Sci. 987, 51–5913. Sims, T. N., and Dustin, M. L. (2002) Immunol. Rev. 186, 100–11714. Purbhoo, M. A., Irvine, D. J., Huppa, J. B., and Davis, M. M. (2004) Nat.

Immunol. 5, 524–53015. Huppa, J. B., Gleimer,M., Sumen, C., and Davis, M.M. (2003)Nat. Immu-

nol. 4, 749–75516. Dustin, M. L., and Springer, T. A. (1989) Nature 341, 619–62417. Kinashi, T. (2005) Nat. Rev. Immunol. 5, 546–55918. Kane, L. P., Lin, J., andWeiss, A. (2000)Curr. Opin. Immunol. 12, 242–24919. Billadeau, D. D., Nolz, J. C., and Gomez, T. S. (2007)Nat. Rev. Immunol. 7,

131–14320. Kucik, D. F., Dustin, M. L., Miller, J. M., and Brown, E. J. (1996) J. Clin.

Invest. 97, 2139–2144

21. Samelson, L. E. (2002) Annu. Rev. Immunol. 20, 371–39422. Mueller, K. L., Daniels, M. A., Felthauser, A., Kao, C., Jameson, S. C., and

Shimizu, Y. (2004) J. Immunol. 173, 2222–222623. Mazerolles, F., Barbat, C., Meloche, S., Gratton, S., Soula, M., Fagard, R.,

Fischer, S., Hivroz, C., Bernier, J., and Sekaly, R. P. (1994) J. Immunol. 152,5670–5679

24. Woods, M. L., Kivens, W. J., Adelsman, M. A., Qiu, Y., August, A., andShimizu, Y. (2001) EMBO J. 20, 1232–1244

25. Goda, S.,Quale, A.C.,Woods,M. L., Felthauser, A., and Shimizu, Y. (2004)J. Immunol. 172, 5379–5387

26. Fagerholm, S., Hilden, T. J., and Gahmberg, C. G. (2002) Eur. J. Immunol.32, 1670–1678

27. Perez, O. D., Mitchell, D., Jager, G. C., South, S., Murriel, C., McBride, J.,Herzenberg, L. A., Kinoshita, S., and Nolan, G. P. (2003)Nat. Immunol. 4,1083–1092

28. Mor, A., Campi, G., Du, G., Zheng, Y., Foster, D. A., Dustin, M. L., andPhilips, M. R. (2007) Nat. Cell Biol. 9, 713–719

29. Smits, H. H., de Jong, E. C., Schuitemaker, J. H., Geijtenbeek, T. B., van,Kooyk, Y., Kapsenberg,M. L., andWierenga, E. A. (2002) J. Immunol. 168,1710–1716

30. Leitinger, B., and Hogg, N. (2002) J. Cell Sci. 115, 963–97231. Marwali, M. R., MacLeod, M. A., Muzia, D. N., and Takei, F. (2004) J. Im-

munol. 173, 2960–296732. Kuhn, J. R., and Poenie, M. (2002) Immunity 16, 111–12133. Sancho, D., Vicente-Manzanares, M., Mittelbrunn, M., Montoya, M. C.,

Gordon-Alonso, M., Serrador, J. M., and Sanchez-Madrid, F. (2002) Im-munol. Rev. 189, 84–97

34. Stinchcombe, J. C., Majorovits, E., Bossi, G., Fuller, S., and Griffiths, G. M.(2006) Nature 443, 462–465

35. Chen, X., Trivedi, P. P., Ge, B., Krzewski, K., and Strominger, J. L. (2007)Proc. Natl. Acad. Sci. U.S.A. 104, 6329–6334

36. Endl, E., Steinbach, P., Knuchel, R., and Hofstadter, F. (1997) Cytometry.29, 233–241

37. Gerdes, J., Lemke, H., Baisch, H.,Wacker, H. H., Schwab, U., and Stein, H.(1984) J. Immunol. 133, 1710–1715

38. Davis, M. M., Krogsgaard, M., Huppa, J. B., Sumen, C., Purbhoo, M. A.,Irvine, D. J., Wu, L. C., and Ehrlich, L. (2003) Annu. Rev. Biochem. 72,717–742

39. Carman, C. V., and Springer, T. A. (2003) Curr. Opin. Cell Biol. 15,547–556

40. Hogg, N., Smith, A., McDowall, A., Giles, K., Stanley, P., Laschinger, M.,and Henderson, R. (2004) Immunol. Lett. 92, 51–54

41. Hynes, R. O. (2003) Science 300, 755–75642. Cairo, C. W., Mirchev, R., and Golan, D. E. (2006) Immunity 25, 297–30843. Bossi, G., Bottone, M. G., and Pellicciari, C. (1998) Eur. J. Histochem. 42,

277–28544. Suzuki, J., Isobe, M., Izawa, A., Takahashi, W., Yamazaki, S., Okubo, Y.,

Amano, J., and Sekiguchi, M. (1999) Transpl. Immunol. 7, 65–7245. Katagiri, K., Maeda, A., Shimonaka, M., and Kinashi, T. (2003) Nat. Im-

munol. 4, 741–748

LFA-1 Regulates Erk1/2 Signal Pathway in CD8� Cells

21010 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 31 • JULY 31, 2009

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from

Dan Li, Jeffrey J. Molldrem and Qing MaLFA-1-mediated Erk1/2 Signal Pathways

T Cell Activation via T Cell Receptor-mediated and+LFA-1 Regulates CD8

doi: 10.1074/jbc.M109.002865 originally published online May 29, 20092009, 284:21001-21010.J. Biol. Chem. 

  10.1074/jbc.M109.002865Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2009/06/02/M109.002865.DC1

  http://www.jbc.org/content/284/31/21001.full.html#ref-list-1

This article cites 45 references, 15 of which can be accessed free at

by guest on February 9, 2018http://w

ww

.jbc.org/D

ownloaded from