IMMUNOBIOLOGY

Activation-induced expression of CD137 permits detection, isolation, andexpansion of the full repertoire of CD8� T cells responding to antigenwithout requiring knowledge of epitope specificitiesMatthias Wolfl,1 Jurgen Kuball,1 William Y. Ho,1 Hieu Nguyen,1 Thomas J. Manley,1

Marie Bleakley,1 and Philip D. Greenberg1,2

1Fred Hutchinson Cancer Research Center, Program in Immunology, Seattle, WA; 2Department of Immunology,University of Washington School of Medicine, Seattle

CD137 is a member of the TNFR-familywith costimulatory function. Here we showthat it also has many favorable character-istics as a surrogate marker for antigen-specific activation of human CD8� T cells.Although undetectable on unstimulatedCD8� T cells, it is uniformly up-regulated24 hours after stimulation on virtually allresponding cells regardless of differentia-tion stage or profile of cytokine secretion,which circumvents limitations of currentsurrogate markers for defining the reper-toire of responding cells based on only

individual functions. Antibody-labeled re-sponding CD137� cells can be easily andefficiently isolated by flow sorting or mag-netic beads to substantially enrich anti-gen-specific T cells. To test this approachfor epitope discovery, we examined invitro priming of naive T cells from healthydonors to Wilms tumor antigen 1 (WT1), aprotein overexpressed in various malig-nancies. Two overlapping pentadecam-ers were identified as immunogenic, andfurther analysis defined WT1(286–293) asthe minimal amino acid sequence and

HLA-Cw07 as the HLA restriction ele-ment. In conclusion, this approach ap-pears to be an efficient and sensitive invitro technique to rapidly identify andisolate antigen-specific CD8� T cellspresent at low frequencies and displayingheterogeneous functional profiles, anddoes not require prior knowledge of thespecific epitopes recognized or the HLA-restricting elements. (Blood. 2007;110:201-210)

© 2007 by The American Society of Hematology

Introduction

Methods to assess antigen-specific CD8 T cells are essential forunderstanding cellular immunity under normal and pathologicconditions. Peptide/MHC (pMHC) multimers1 accurately detectantigen-specific T cells, but the knowledge of the immunogenicpeptide, its restricting HLA allele, and the availability of thereagent are limiting and rarely sufficient to characterize thebreadth of responding cells. Evaluating the repertoire of re-sponses has required identifying—and often subsequently isolat-ing—antigen-specific T cells using assays that focus on specificfunctions such as production of a particular cytokine ordegranulation in response to antigen.2,3 However, such func-tional assays often reflect only a subset of specifically activatedcells, as the in vivo CD8� T-cell response is composed offunctionally diverse cells that cannot be identified using a singlecytokine or the detection of degranulation.4,5 For example,different T-cell subpopulations, such as naive T cells, centralmemory T cells, or T cells primed under Th1- or Th2-likeconditions, show very heterogeneous cytokine profiles, and cantherefore be tracked only functionally using multicytokine flowcytometry.6 A surface molecule that is uniformly up-regulated inresponse to antigen would therefore facilitate the assessment ofthe complete antigen-specific response. Favorable characteris-tics of such a surrogate marker would include specific surfaceexpression after activation over a transient but sufficientlyprolonged time period to allow reliable detection, and absent

expression if unstimulated and during resting phases. Such anexpression pattern would facilitate not only the identificationbut also the selection of viable, activated T cells using fluores-cence-activated cell sorting or magnetic beads bound to thespecific antibody.

CD137 (4–1BB) was originally identified as a moleculeexpressed on activated mouse and human CD8� and CD4�

T cells.7-9 It is a member of the TNFR family and mediatescostimulatory and antiapoptotic functions, promoting T-cellproliferation and T-cell survival.10,11 CD137 has been reportedto be up-regulated—depending on the T-cell stimulus—from12 hours to up to 5 days after stimulation.9,10,12 Antibody-mediated blockade of the CD137/CD137L interaction has beenshown to increase survival of murine allografts,13 and depletionof human CD137� donor-derived T cells after stimulation withallogeneic recipient cells in vitro has recently been described asan approach to reduce alloreactivity.14 Here we demonstrate thatCD137 has many characteristics required for efficient positiveselection of specifically activated CD8� T cells with a heteroge-neous functional profile, including memory and naive responsesto viral-associated (cytomegalovirus [CMV] and hepatitis Cvirus [HCV] related) and tumor-associated self-antigens such asMART1/Melan-A15 and WT1, a transcription factor overex-pressed in many malignancies and a potential target for immuno-therapy.16,17 Furthermore, this approach for positive selection of

Submitted November 8, 2006; accepted March 2, 2007. Prepublished onlineas Blood First Edition paper, March 19, 2007; DOI 10.1182/blood-2006-11-056168.

The publication costs of this article were defrayed in part by page charge

payment. Therefore, and solely to indicate this fact, this article is herebymarked ‘‘advertisement’’ in accordance with 18 USC section 1734.

© 2007 by The American Society of Hematology

201BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

responding cells appears useful for identifying and validatingT-cell responses to unknown antigens or epitopes, as evidencedby our studies characterizing a novel epitope in WT1.

Materials and methods

Peripheral blood mononuclear cells (PBMCs) from healthy donors wereharvested by leukopheresis after informed consent in accordance with theDeclaration of Helsinki. Institutional review board (IRB) approval for thesestudies was granted by the IRB of the Fred Hutchinson Cancer ResearchCenter. T-cell cultures were maintained in RPMI 1640 medium supple-mented with 12.5 mM HEPES, 4 mM L-glutamine, 100 U/mL penicillin,and 100 �g/mL streptomycin (Invitrogen, Carlsbad, CA), 50�M �-mercap-toethanol (Sigma, St Louis, MO), and 10% human serum. Melanoma celllines A375 (gift from S. Rosenberg, NCI, Bethesda, MD) and Mel526 (giftfrom M. Lotze, University of Pittsburgh, PA) were maintained in RPMI1640 with 25 mM HEPES, 4 mM L-glutamine, 50 U/mL penicillin,50 mg/mL streptomycin, 10 mM sodium pyruvate, 1 mM nonessentialamino acids, and 10% FBS (HyClone, Logan, UT). Both lines express theHLA-A2 allele, but only Mel526 expresses the MART-1 Ag. The T2 cellline is a TAP-deficient T-cell/B-cell hybrid expressing the HLA-A2 allele.

Peptides

The HLA-A*0201–restricted peptides representing defined epitopes fromMelan-A(26-35L) (ELAGIGILTV), WT1(126-134) (RMFPNAPYL), HIV-1/gag/p17(77-85) (SLYNTVATL), CMV/pp65(495-503) (NLVPMVATV), and CMV/pp65(417-426) (TPRVTGGGAM) were synthesized by Synpep (Dublin, CA).HCV/NS3(1406-1415) (KLVALGINAV) was synthesized by Proimmune (Ox-ford, United Kingdom). A peptide library spanning WT1, consisting of15-mers overlapping by 11, was purchased from JPT (Berlin, Germany).

Induction of T-cell lines

T-cell lines were generated as previously described with minor modifica-tions.18 Briefly dendritic cells (DCs) were derived from adherent monocytesby culture for 5 days in Cellgenix DC Medium (Cellgenix, Freiburg,Germany) supplemented with 1% human serum, 800 IU/mL GM-CSF(Berlex, Seattle, WA), and 1000 IU/mL IL-4 (R&D Systems, Minneapolis,MN). On day 3, 1.5 mL fresh medium, supplemented with 1600 IU/mLGM-CSF and 1000 IU/mL IL4, was added to each well. On day 5, cellswere harvested and matured for 24 hours in fresh DC medium supple-mented with a cytokine cocktail of 800 IU/mL GM-CSF, 1000 IU/mL IL-4,10 ng/mL LPS (derived from E coli 055:B5; Sigma), and 50 IU/mL IFN�(Intermune, Brisbane, CA). During the maturation phase, DCs were pulsedwith 10 �g/mL peptide. For experiments using the peptide library, eachindividual peptide was at a concentration of 1.8 �g/mL.

Naive CD8� T cells were obtained by depleting CD45RO� cells fromPBMCs using anti-CD45RO microbeads (Miltenyi Biotec, Auburn, CA),followed by positive selection using anti-CD8 microbeads. This approachresulted in more than 95% pure CD45RO�CD8� populations.

Mature, peptide-pulsed DCs were washed and incubated with CD8�

cells at a ratio of 1:2-1:10 in complete T-cell medium supplemented with100 �M 1-methyltryptophan (Sigma) and each was distributed in 96-wellv-bottom plates or 24-well-plates (2 � 105 to 5 � 105 T cells/well, respec-tively). On day 4, IL-7 and IL-15 (final concentration (f.c.) 5 ng/mL; R&DSystems) were added with medium. Cell lines were supplemented every 2to 3 days with medium and cytokines.

For restimulation, autologous PBMCs were pulsed with peptide(10 �g/mL) for 2 hours and irradiated (30 Gy). Each T-cell line washarvested, resuspended in medium without cytokines, and mixed with1 � 106 PBMCs in a 6-well plate. Twenty-four hours later, 2 mL mediumwas added containing cytokines (f.c. 50 IU/mL IL-2, 5 ng/mL IL-7, and5 ng/mL IL-15). T-cell cloning was performed as described recently.18

Immunophenotyping and pMHC-multimer staining

Antibodies used for immunophenotyping include anti–CD8-FITC or -PE,anti–CD69-PE, anti–CD25-PE or -FITC, anti–CD137-PE or -APC, anti–

CD134-PE, anti–HLA-DR-PE, and anti–CD38-PE (all from BD Pharmin-gen, San Diego, CA). pMHC-multimer staining was performed usingAPC-labeled pMHC multimers (Proimmune) and incubating the cells for20 minutes at room temperature (RT), followed by 20-minute incubation atRT with anti-CD8 antibody. Dead cells were excluded using 7-AAD (BDPharmingen).

Intracellular cytokine staining and CD107a mobilizationshift assay

The assay was performed as described.3,19 Briefly, 2 � 105 T cells werestimulated for 5 hours in 96-well plates using 4 � 105 peptide-pulsed T2cells or DCs. CD107a-FITC (BD) and 10 �g/mL brefeldin A (Sigma) wereadded at the beginning of the stimulation period. After 5 hours, cells werestained for CD107a and costained for CD8. Cells were fixed, permeabil-ized, and stained with antibodies against IFN�, TNF�, or IL-2 (BD), usingFIX/PERM and PERM/Wash solution (BD).

CD137 enrichment assay, IFN�-secretion assay, and chromiumrelease assay

T-cell lines were generated from naive T cells as described in “Introduc-tion.” Twenty-four hours after restimulation, T cells were washed andstained for 20 minutes on ice with anti–CD137-APC (5 �L/100 �L) (BD).After washing, the cells were incubated with anti–APC-microbeads (Milte-nyi Biotec) and selected using MS or LS columns.

The IFN�-secretion assay (Miltenyi Biotec) was performed as de-scribed by the manufacturer. The peptide stimulus was the same as forCD137 enrichment, with a stimulation period of 4 hours to detect peakIFN� responses. The cells were allowed to secrete IFN� for 45 minutes, andthen cooled in the appropriate volume of cold buffer. All further procedureswere performed on ice.

Chromium release assays were performed as described elsewhere.18

Statistics

Pearson correlation (2-tailed) for data derived from 2 methods wascalculated using GraphPadPrism software (GraphPad, San Diego, CA).For each analysis, the P value and the coefficient of determination (r2)are indicated.

Results

To identify surface markers that could serve as surrogate markersfor functional activation, a T-cell clone specific for the heterocliticHLA-A*0201–restricted epitope of MelanA/MART1(26–35L) thathad entered a resting phase at the end of a 14-day expansion cyclewas stimulated with peptide-pulsed T2 cells as APCs. Twenty-fourhours after stimulation, CD69, CD25, CD137, CD38, and HLA-DRwere all up-regulated (Figure 1A). Among these, CD69, CD25, andCD137 exhibited a large enough shift in fluorescence intensity topermit separation of activated cells from unstimulated cells. CD137expression was increased after 5 hours, peaked by 24 hours, andthen gradually declined during the period from 48 to 72 hours(Figure 1B). Among the other markers, only CD25 showed similarkinetics to CD137, whereas expression of CD69, CD38, andHLA-DR was more variable, usually peaking after 6 hours butthen quickly returning to baseline levels. CD134 was stillnegative at 24 hours, but increased slightly by 48 hours. Amongthe 3 markers that were up-regulated sufficiently for specificdetection—CD25, CD137, and CD69—only CD137 was notexpressed on unstimulated, resting T cells, which is essential ifthe marker is to be used for a bead-based selection method.Similar results were obtained for 5 different T-cell clones withvarious specificities (data not shown).

202 WOLFL et al BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

As CD25, CD137, and CD69 were most informative, theirexpression was next analyzed in a Melan-A(26–35L)–specific T-cellline to assess baseline expression and up-regulation in a polyclonalpopulation, generated by 2 rounds of stimulation. Twenty-fourhours after the third stimulation the cells were stained usingPE-labeled mAbs for the respective markers and costained withanti-CD8-antibody and the pMHC multimer. As expected, TCRdown-regulation in response to antigen hampered pMHC-multimerstaining, with fewer multimer� cells and lower intensity stainingdetected at 24 hours compared with before stimulation (Figure 1C),making this unreliable for tracking antigen-specific cells afteractivation. All 3 markers were strongly up-regulated on a largefraction of cells. However, similar to what had been observed withclones, baseline expression was minimal for CD137, whereas low

or intermediate expression for CD69 and CD25, respectively, wasobserved in unstimulated cells (Figure 1C). In a parallel set ofsamples, triple staining for CD25, CD69, and CD137 was per-formed. CD137� cells were nearly uniformly triple positive (96%),whereas gating on CD69� or CD25� cells defined more heteroge-neous populations with 18% and 17%, respectively, negative forCD137, which may represent nonspecifically stimulated cells inculture expressing the observed higher baseline levels of CD69 andCD25 (data not shown). Of note antigen-independent up-regulationof CD137 in response to IL-15, as has been reported with highdoses (100 ng/mL),20 was not observed in the culture systems usedfor this study, as a lower more physiologic dose (5 ng/mL) of IL-15was routinely used. Thus, CD137 appeared to have the mostfavorable expression profile among the markers tested for identify-ing specifically activated CD8� T cells.

CD137 expression correlates with functional activation of CD8T-cell clones and lines

As CD137 up-regulation is a function of T-cell activation, thethreshold for inducing CD137 expression was compared with IFN�and TNF� production after stimulation with titrated amounts ofthe cognate antigen. Two T-cell clones, clone 88 specific forMelan-A(26-35L) and clone 10 specific for WT1(126–134), were stimu-lated with T2-cells pulsed with titrated amounts of peptide. Foroptimal results, these functional assays require different incubationtimes—therefore samples were set-up in parallel and cytokineproduction assessed after 5 hours and CD137 expression deter-mined after 24 hours of incubation. The cells showed comparableactivation thresholds for cytokine production and CD137 ex-pression, with similar peptide concentrations necessary to attainhalf-maximal activation (Clone 88 with an EC50 [M] forCD137 � 2.5 � 10�9, IFN� � 3.7 � 10�9, and TNF� � 1.5 � 10�9;Clone 10 with an EC50 [M] for CD137 � 9.5 � 10�9,IFN� � 2.0 � 10�8, and TNF� � 9.2 � 10�9) (Figure 2A). The corre-lation between cytokine production and CD137 expression was highlysignificant (Clone 88 with [P/r2] for CD137 vs IFN� � P .001/0.99and CD137 vs TNF� � P .001/0.98; Clone 10 with [P/r2] for CD137vs IFN� � P .001/0.97 and CD137 vs TNF� � P .001/0.98).

The HLA-A*0201–restricted heteroclitic Melan-A/MART1(26-35L)

epitope is unique in that a fairly high precursor frequency of naiveantigen-specific T cells in healthy humans can often be found.15

Therefore, to examine a prototypic in vitro response with a more typicalprecursor frequency, responses to an A*0201-restricted epitope derivedfrom the nonstructural protein 3 of the hepatitis C virus (NS3(1406–1415))21

were studied in cell lines derived from healthy, HCV seronegativehumans. To compare CD137 expression with other standard assays usedfor identification of activated T cells in this setting of lower frequencyresponses, cytokine flow cytometry (CFC) and the CD107a-mobilization shift assays to assess degranulation3 were per-formed in parallel. Excellent correlations of CD137 expressionwere found with CFC data (CD137 vs IFN� [P;r2] � .002;0.92and CD137 vs TNF� [P;r2] � .001;0.96), as well as with theCD107a mobilization shift assay (CD137 vs CD107a[P;r2] � .007;0.87) and pMHC multimer staining (CD137 vspMHC multimer [P;r2] � .007;0.86) (Figure 2B).

We next examined if CD137 could serve as a surrogatemarker for antigen-specific activation to assess the response ofmemory CD8� T cells in PBMCs ex vivo. PBMCs from7 HCMV-seropositive, HLA-A*0201–positive donors werethawed, rested overnight, and then stimulated to comparefunctional responses to the immunodominant epitope of

CD134

HLA-DR

CD69 CD38

CD25 CD137

5 24 48 72 96

1

5

10

15

20

25

CD69

HLA-DRCD38

CD25CD137

CD134

hours after stimulation

x fo

ld i

ncr

ease

of

MF

I

0.2 0.5

32 67

0.5 0.5

32 67

4.0 2.4

30 63

17 45

16 22

60 27

13 0.1

0.7 0.3

72 27

61 23

14 1.7

47 25

26 2.0

MHC-multimer

Isot

ype

CD

25C

D69

CD

137

unstimulated stimulated

A

B

C

Figure 1. Expression level and kinetics of activation markers after antigen-specific activation. To evaluate a panel of activation markers, a T-cell clone specificfor the HLA-A*0201–restricted epitope of Melan-A(26-35L) was stimulated with peptide-pulsed T2 cells, and assessed for (A) antigen-specific up-regulation of denotedmarkers 24 hours after stimulation (thick lines) compared with control peptide (filled)or isotype-matched antibodies (dotted line); and (B) expression kinetics of the panelof markers in the same experiment (x-fold increase: MFI of the specifically stimulatedsample/MFI of the sample stimulated with the irrelevant control peptide).(C) A polyclonal Melan-A–specific T-cell line generated by 2 in vitro stimulations wasassessed for baseline expression and specific up-regulation 24 hours after stimula-tion of the 3 most informative activation markers (numbers indicate the percentage ofviable CD8� T cells in each quadrant).

CD137 IDENTIFIES CD8 T CELLS RESPONDING TO ANTIGEN 203BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

pp65(495–503). Again, CD137 expression correlated with cytokineproduction ([P;r2] for IFN� � .001;0.9 and TNF� � .001;0.92),degranulation ([P;r2] for CD107a � .002;0.87) and pMHC multi-mer staining ([P;r2] .012;0.75) (Figure 2C), demonstrating thatCD137 can be used as a surrogate marker to detect direct activationof responding cells in PBMCs. Of note, proliferation of stimulatedantigen-specific cells had not commenced before the time ofCD137 staining, as determined by CFSE labeling (data not shown),affirming that the results reflect the number of antigen-specific cellscapable of responding in PBMCs detected by each assay.

Viable CD137� T cells can be easily isolated and areantigen-specific

Without knowledge of the recognized epitope and an availabletetramer, the isolation and expansion of T cells specific for aselected target or antigen for either analysis or use in adoptivetherapy can be difficult. In initial experiments, we were unsuccess-ful in generating robust manipulable responses to a panel ofcandidate peptides derived from the weakly immunogenic WT1

antigen, using the IFN�-secretion assay to isolate cells (data notshown). Therefore, as an alternative strategy, we examined selec-tion of cells that demonstrated up-regulation of CD137 afterantigen stimulation for enrichment of responding T cells fromT cell cultures containing an initially low frequency of antigen-specific T cells. Before assessing this technique for epitopediscovery of unknown antigens, as described later for WT1, wewanted to validate the approach using a model foreign antigenknown to induce responses. A polyclonal T cell response wasgenerated against the HCV/NS3(1406-1415) epitope by stimulation ofnaive CD8� T cells from an HCV-seronegative donor withpeptide-pulsed DCs, and cultured for 1 week in medium containingIL-7 and IL-15 as previously described.18 CD45RO-CD8� cellswere routinely used as the naive starting population, but a similarT-cell response against this epitope could also be induced usingCD62L�CD45RO�CD8� T cells or CD45RA�CCR7�CD8�

T cells. In contrast, HCV/NS3(1406-1415)-specific T cells could notbe generated from CD45RO� T cells (data not shown). Theprimed cells were restimulated and enriched for CD137� cells

0 -12-11 -10 -9 -8 -7 -60

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 -12-11-10 -9 -8 -7 -6 -50

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 -12-11 -10 -9 -8 -7 -60

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 -12-11 -10 -9 -8 -7 -60

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 -12-11-10 -9 -8 -7 -6 -50

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 -12-11-10 -9 -8 -7 -6 -50

25

50

75

100

log peptide concentration [M]

% p

os

itiv

e c

ell

s

0 10 20 30 40 500

10

20

30

40

50

0.0023;0.92

% CD137+

% I

FN

-�+

0 10 20 30 40 500

10

20

30

40

50

0.0005;0.96

% CD137+

% T

NF

-�+

0 10 20 30 40 500

10

20

30

40

50

0.0065;0.87

% CD137+%

CD

10

7a

+0 10 20 30 40 50

0

10

20

30

40

50

% CD137+

%p

MH

C m

ult

ime

r 0.0074; 0.86

0 10 20 30 40 500

10

20

30

40

50

0.0023;0.92

% CD137+

-

0 10 20 30 40 500

10

20

30

40

50

0.0005;0.96

% CD137+

% T

NF

-

0 10 20 30 40 500

10

20

30

40

50

0.0065;0.87

% CD137+%

CD

10

7a

+0 10 20 30 40 50

0

10

20

30

40

50

% CD137+

%p

MH

C m

ult

ime

r

0 10 20 30 40 500

10

20

30

40

50

0.0023;0.92

% CD137+

-

0 10 20 30 40 500

10

20

30

40

50

0.0023;0.92

% CD137+

-

0 10 20 30 40 500

10

20

30

40

50

0.0005;0.96

% CD137+

% T

NF

-

0 10 20 30 40 500

10

20

30

40

50

0.0005;0.96

% CD137+

% T

NF

-

0 10 20 30 40 500

10

20

30

40

50

0.0065;0.87

% CD137+%

CD

10

7a

+

0 10 20 30 40 500

10

20

30

40

50

0.0065;0.87

% CD137+%

CD

10

7a

+0 10 20 30 40 50

0

10

20

30

40

50

% CD137+

%p

MH

C m

ult

ime

r

0 10 20 30 40 500

10

20

30

40

50

% CD137+

%p

MH

C m

ult

ime

r 0.0074; 0.86

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0012;0.9

% CD137+

% I

FN

-�+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0006;0.92

% CD137+

% T

NF

-�+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0022;0.87

% CD137+

% C

D1

07

a+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

% CD137+

% p

MH

C m

ult

ime

r+ 0.0117; 0.75

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0012;0.9

% CD137+

% I

FN

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0006;0.92

% CD137+

% T

NF

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0022;0.87

% CD137+

% C

D1

07

a+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0012;0.9

% CD137+

% I

FN

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0012;0.9

% CD137+

% I

FN

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0006;0.92

% CD137+

% T

NF

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0006;0.92

% CD137+

% T

NF

-

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0022;0.87

% CD137+

% C

D1

07

a+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

0.0022;0.87

% CD137+

% C

D1

07

a+

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

% CD137+

% p

MH

C m

ult

ime

r+ 0.0117; 0.75

0 1 2 3 4 5 6 7 80

1

2

3

4

5

6

7

8

% CD137+

% p

MH

C m

ult

ime

r+ 0.0117; 0.75

Clone88/Melan-A(26-35L) Clone10/WT1(126-134)A

B

C

Figure 2. CD137 up-regulation as a surrogate marker for specifically activated CD8� T cells in vitro and ex vivo. The activation threshold for several markers ofresponse was compared in responding T-cell clones (A), T-cell lines (B), or memory T cells directly ex vivo (C). (A) Two HLA-A*0201–restricted T-cell clones, specific for eitherMelan-A(26-35L) or WT1(126-134), were stimulated with titrating amounts of peptides. IFN� (squares) and TNF� (triangles) production was assessed after 5 hours of stimulation inthe presence of brefeldin A, and CD137 expression (circles) was determined 24 hours after stimulation (note: these reflect optimal time points for detecting each of theseevents). (B) Six separate T-cell lines specific for the NS3(1406-1415)-epitope of HCV were generated, and assessed for cytokine production and degranulation after 5 hours ofstimulation, and CD137 expression after 24 hours and pMHC-multimer staining of the unstimulated sample. Pearson correlation (2-tailed) was calculated, and the P value andthe coefficient of determination (r2) appear in each plot. (C) PBMCs from 7 CMV�/HLA-A*0201� donors were thawed, rested overnight, and analyzed for pMHC multimerstaining, cytokine production, and degranulation after 5 hours of stimulation and CD137 expression at 24 hours. P value; r2 appear in each plot.

204 WOLFL et al BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

24 hours later using magnetic beads, and placed back incytokine-containing medium for 8 days to allow expansion andre-expression of the TCR before being analyzed using pMHCmultimers (Figure 3A). The frequency of HCV/NS3(1406-1415)-specific T cells in fresh PBMCs from seronegative, healthydonors, as assessed by pMHC-multimer staining, is below thelimits of detection (not shown), but a small percentage (0.39%)of pMHC-multimer� cells was detectable 8 days after primarystimulation (Figure 3Bi). After a second stimulation, no signifi-cant enrichment of this small population of reactive cells wasdetectable (Figure 3Biii), which is not uncommon for low-

frequency responses presumably due to cell death of some of thespecific T cells and cytokine-mediated expansion of nonspecificbystander T cells. However more reactive cells were presentthan if the line remained in culture and was not restimulatedwith antigen (Figure 3Bii; 0.34% vs 0.05%). In contrast, ifenriched after restimulation and expanded for 1 week, 27% ofthe cells were pMHC-multimer� T cells (Figure 3Biv), represent-ing a 79-fold enrichment compared with cells not enriched and a9-fold absolute expansion in antigen-specific cell numbers(Table 1). Adding the antibody to the cells, without enrichment,did not have a measurable stimulatory effect (data not shown).

CD137-

CD137+

8 daysIL2/7/15

pMHC multimer staining

Enrichment

24h

Restimulation with irradiated, peptide-pulsed PBMC

Primary stimulation with peptide-pulsedmature DC and naive CD8+T-cells

8 daysIL7/IL15

0.39% 0.05%

27%0.34%

i ii

iii iv

CD

8

pMHC multimer

0.4 19.6 0.1 0.0

78.3 1.8 0.1

0.3 9.7 0.1 0.0

78.4 11.6 0.0

Interferon-gamma

TNF-

alp

ha

IL2

stim

ulat

ed/n

ot enr

iche

d

Stimul

ated

/CD13

7 en

riche

d

0.1

1

10

100

% M

HC

mu

ltim

er

+ c

ells

stim

ulat

ed/n

ot enr

iche

d

Stimul

ated

/CD13

7 en

riche

d

0.1

1

10

100

% M

HC

mu

ltim

er

+ c

ells

0.1

1

10

100

% M

HC

mu

ltim

er

+ c

ells

A

B

C

D

Figure 3. Use of CD137 expression for enrichment of antigen-specific T cells. (A) Experimental outline: Naive CD45RO�CD8� T cells were stimulated with peptide-pulsedDCs and cultured for 1 week with IL-7 and IL-15 added on day 4. On day 8, the cells were restimulated with irradiated, autologous, peptide-pulsed PBMCs. Twenty-four hourslater, CD137� cells were isolated with magnetic beads and cultured for an additional 8 days in medium containing IL-2, IL-7, and IL-15 to allow both expansion andre-expression of the TCR. The cells were then analyzed for specificity using pMHC multimers. (B) Enrichment of an HCV/NS3(1406-1415)-specific T-cell line by selection ofCD137� cells. Eight days after primary stimulation of naive T cells, an aliquot was stained with an NS3(1406-1415)/MHC multimer to determine the frequency of antigen-specificcells before restimulation and/or enrichment (i). These cells were then split and either cultured in IL-2–, IL-7–, and IL-15–containing medium without any restimulation (ii),restimulated and cultured but not enriched (iii), or restimulated and enriched for CD137� cells 24 hours later and then cultured for an additional 8 days (iv), and then stained withanti-CD8 antibody and the pMHC multimer. (C) Parallel to the pMHC-multimer staining, an aliquot of the enriched T-cell population was also restimulated with peptide-pulsedT2 cells and cytokine production was assessed after 5 hours (left). T cells incubated with unloaded T2 cells served as control (right). (D) Nine different HCV/NS3-specific T-celllines were generated and enriched 24 hours after the second stimulation as above, and the number of pMHC multimer� T cells assessed 8 days later and compared with linesnot enriched (2-tailed, paired t test: P � .002).

Table 1. CD8� T cells, selected based on their expression of CD137, expand after in vitro culture

Exp.no. Epitope

% multimer�

beforeselection

Total no. ofselected

cells, �105

% multimer� cells1 wk afterselection

Absolute no.of cells after1 wk, �106

Absolute no. ofantigen-specific

cells,* �106Expansion factor,

x-fold

1 HCV(1406-1415) 0.39 1.1 27 3.8 1.0 9.3

2 HCV(1406-1415) 1.6 6.0 61 8.1 4.9 8.2

3 WT1(126-134) 2.4 2.2 49 2.3 1.1 5.1

4 MelanA(26-35L) 25 4.5 72 4.5 3.2 7.2

5 MelanA(26-35L) 8 1.4 87 2.2 1.9 13.7

*The absolute number of antigen-specific cells was calculated by (the absolute number of cells � percentage of MHC multimer� cells)/100.

CD137 IDENTIFIES CD8 T CELLS RESPONDING TO ANTIGEN 205BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

To demonstrate that the enriched T cells were functional,intracellular cytokine staining in response to peptide wasperformed. More than 20% of the CD8 T cells (79% of the MHCmultimer� cells) produced IFN� and most of them also pro-duced TNF�. Almost 10% produced IL-2, representing 36% ofthe MHC multimer� population (Figure 3C). The enrichmentprocedure was repeated for 9 individual T-cell lines andenrichment resulted in achieving frequencies of 3% to 33%(mean, 16.15%) (Figure 3D). The average enrichment, calcu-lated from the pooled data of these experiments, was 74-foldover the stimulated but unenriched control. To assess reproduc-ibility of the enrichment procedure, one T-cell line was gener-ated as described above, restimulated and then split in 5 aliquotsprocessed for enrichment in parallel. The mean percentage ofpMHC multimer� cells 1 week after enrichment was 8.6% witha standard error of 0.92% (data not shown). The procedure wasalso applicable for different peptide antigens (Melan-A(26-35L),HIV-1 gag p17(77-85)) and for T-cells lines after various2-4 roundsof stimulation (data not shown).

Due to the exceedingly low precursor frequency of antigen-specific T cells in the naive repertoire, multiple rounds of restimu-lation are often required until antigen-specific T cells can besuccessfully and efficiently expanded on a clonal level. Becauseenrichment for CD137� cells in bulk T-cell populations wasfeasible after only 2 stimulations (Figure 3B), we asked whetherthe approach would also allow for clonal expansion of T cells afteronly 2 rounds of in vitro stimulation. Purified naive CD8� T cellswere stimulated with the HCV/NS3 epitope and, 24 hours after thesecond stimulation, T cells were cloned from either the unenrichedor from the CD137-enriched group. At 30 days from the firststimulation of the T cells (20 days after cloning), the growingclones were screened using pMHC multimers. In the CD137�

group, 250 of 480 plated wells contained growing cells. Sixty-seven percent of these wells contained clones that were HCV/NS3(1406-1415) specific (overall specific cloning efficiency, 35%). Incontrast, only 75 of 480 wells from the unenriched control groupshowed cell growth, with only 15% of these wells containingclones that were pMHC multimer� (overall specific cloningefficiency, 2.3%) (Table 2).

CD137 is a more uniform marker of antigen-specific activationthan monitoring production of only a single cytokine

Methods based on cytokine secretion after activation are often usedto enrich antigen-specific CD8� cells, and we compared the utility

of the IFN�-secretion method with the CD137-enrichment methodfor selecting responding T cells derived from memory or naivepopulations. Both assays were compared for direct ex vivoenrichment from a starting population of CMV-specific memoryT cells, which are known to secrete high levels of IFN�.22 PBMCsfrom a CMV� patient were stimulated with the HLA-B7–restrictedepitope CMVpp65(417-426) and responding cells enriched based oncapture of IFN�-secreting cells at 4 hours or enrichment forCD137� cells at 24 hours after stimulation. Flow cytometriccontrols showed that both procedures resulted in comparable andadequate enrichment (data not shown). One week after the enrich-ment step, the groups were compared by pMHC multimer staining.Both assays resulted in excellent enrichment (IFN� � 76% andCD137 � 96%) (Figure 4Aiii-iv). To determine how effectivelythe 2 methods define the complete antigen-specific population, thenegative fractions were examined for remaining antigen-specificT cells. To ensure complete removal of cells that can be identifiedby either method, the negative fractions (flowthrough) from theabove enrichment procedures were passed over a second columndesigned to deplete any remaining cells attached to beads, and thedepleted fraction cultured and analyzed one week later. After

i ii

iii iv

v vi

1.2% 1.2%

76% 96%

0.2% 0.0%

CD

8

CMVpp65/B7-multimer

CD

8

Melan-A/A2-multimer

0.1% 36%

CD137

38%

INF-TNF-IL2+9%

IFN-TNF+IL2+5%

IFN-TNF+IL2-12%

IFN+TNF-IL2-34%

IFN+TNF+IL2-22%

IFN+TNF+IL2+9%

IFN+TNF-IL2+9%

A B

C

Figure 4. CD137 enrichment is effective with memory and naive T-cell popula-tions. (A) PBMCs from a CMV�/HLA-B7� donor were used to compare the CD137enrichment method versus the IFN�-secretion assay (all plots, except the ex vivostaining, show groups after an additional 8 days of culture): (i) CMVpp65(417-426)-pMHC multimer� cells ex vivo; (ii) stimulation with peptide, no enrichment;(iii) stimulation with peptide and enrichment with IFN�-secretion assay; (iv) stimula-tion with peptide and CD137 enrichment; (v) IFN� group (flowthrough/depleted);(vi) CD137� cells (flowthrough/depleted). (B) PBMCs from a healthy donor werestimulated with Melan-A(26-35L) peptide directly ex vivo for 24 hours and then eithercultured for an additional week in medium containing IL-7 and IL-15 (left), or enrichedfor CD137� cells and cultured for one week (right). The enriched sample was thenrestimulated a second time (day 8 of culture) and assessed for (C) up-regulation ofCD137 at 24 hours after peptide-specific stimulation (left, black line) or stimulationwith irrelevant control peptide (filled). In parallel, the pattern of cytokine responseswas determined by intracellular cytokine staining (right, 5-hour stimulation period, %responsive CD8� cells noted).

Table 2. Cloning of HCV-specific T cells after the second stimulationwith or without CD137 enrichment*

Notenriched

CD137enriched

Total no. of wells 480 480

Percentage of specific T cells prior to cloning 0.11 0.11

Initial cell no./well 4 4

Total no. of wells with positive cell growth (%) 75 (16) 250 (52)

No. of wells tested by pMHC-multimer 55 64

% of multimer� clones/CD8� clone 15 67

% of multimer� cells in parallel bulk culture 12 62

Overall specific cloning efficiency, %* 2.3 35

Cloning was performed 24 hours after the second stimulation in mediumcontaining irradiated allogenous B-LCLs and PBMCs as feeder cells and IL-2 andIL-15. Plates were evaluated by pMHC-multimer staining 20 days after cloning.

*Specific cloning efficiency was determined as the percentage of wells withspecific cell growth in relation to the total number of wells.

206 WOLFL et al BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

depletion of IFN�� cells, a small percentage (0.2%) of pMHCmultimer binding CMV� T cells was detectable (Figure 4Av),suggesting that some cells had either not been labeled sufficientlyby the antibody for selection or were not secreting IFN� at the timeof enrichment. In contrast, no pMHC-multimer� cells were de-tected 1 week after depletion of CD137� cells, suggesting that by24 hours after stimulation CD137 is uniformly up-regulated on allantigen-specific T cells (Figure 5Avi). Similar results were ob-tained when analyzing the negative fractions (flowthrough) afterstimulation of T-cell lines against different specificities without useof an additional depletion column (data not shown).

Unlike the CD8 memory response to CMV in which IFN� is thedominant activation-induced cytokine,22 naive T cells are known tosecret low quantities of IFN�, a function that is acquired later afterdifferentiation into Tc1-like effector cells. It is therefore difficult toisolate naive CD8� T cells using the IFN�-secretion assay as hasbeen described before.23 Indeed, after stimulating naive T cells withthe Melan-A peptide, we were unable to enrich for antigen-specificcells using the IFN�-secretion method (data not shown). Howeverwhen PBMCs were stimulated directly ex vivo with the Melan-Apeptide and CD137� cells isolated 24 hours later and cultured for 1week, 36% of the cells were pMHC multimer� (Figure 4B rightpanel). Stimulation of PBMCs with peptide without enrichment didnot result in the outgrowth in 1 week of a detectable Melan-A–specific population from this donor (Figure 4B left panel), whosenatural frequency of naive Melan-A–specific cells in PBMCs wasbelow the detection limit with pMHC multimers (data not shown).Thus, in contrast to IFN� production, CD137 can be used for earlyisolation of responding CD8� T cells from primary responses.

The functional response of these cells that had been enriched forCD137� cells 24 hours after primary sensitization was assessedafter 1 week in culture by restimulation with peptide-pulsedPBMCs. The percentage of CD137� cells detected 24 hours afterrestimulation correlated well with the percentage of pMHC-multimer� cells present before restimulation (Figure 4C left paneland 4B right panel). However, consistent with the origin of thesecells from a naive repertoire and stimulation under conditionslacking strong Th1/Tc1-polarizing signals, the responding cellsexpressed a heterogeneous cytokine profile, with 26% of theresponding cells making either IL-2, TNF�, or both, but not IFN�(Figure 4C right panel). Thus, selection based on production of asingle cytokine alone, such as IFN�, would have resulted in anunderestimation of this responding population.

Enrichment of CD137� T cells upon antigen stimulationincludes T cells with lytic capacity against antigen� tumor cells

We next wanted to know if the isolated and enriched T-cellpopulation includes T cells that were able to recognize endog-enously processed and presented antigen by tumor cells. AMelan-A–specific T-cell line was generated by 2 rounds of in vitrostimulation, resulting in 8% pMHC-multimer� T cells (Figure5Ai). This line was restimulated with peptide-pulsed PBMCs andCD137� cells were selected 24 hours later (Figure 5Aii). Whenthese cells were further expanded for one week, 87% of the T cellswere Melan-A multimer� (Figure 5Aiii). This line was then testedfor recognition of the HLA-A2�, Melan-A� melanoma cell lineMel526, or the HLA-A2� Melan-A� melanoma cell line A325.Twenty-four hours after coincubation, 57% of the T cells up-regulated CD137 in response to Mel526 (Figure 5Aiv), whereas norelevant up-regulation was observed when incubated with theMelan-A� tumor line A325 (Figure 5Av). Most importantly,specific lysis of the Melan-A� tumor targets was observed only

in the enriched T-cell line, whereas cells saved from theflowthrough after enrichment did not show lytic capacity(Figure 5B). This demonstrates, that the CD137� populationcontains affinity T cells capable of recognizing endogenouslyprocessed and presented antigen.

The CD137 assay as a tool for adoptive immunotherapyand epitope discovery

In the context of malignancy, a selection procedure that does notrequire the exact knowledge of the epitope or the HLA-restrictionelement could be used to identify novel antigens and broaden thespectrum of patients eligible for immunotherapy trials.24 WT1,a pro-oncogenic protein overexpressed in hematologic and onco-logic malignancies,16 is being evaluated as a target protein forimmunotherapy.17 As only few epitopes are defined so far,25-27 weexamined if novel WT1-specific epitopes could be identified byselecting for CD137� T cells responding to stimulation with apeptide library, consisting of 15-mers spanning the complete WT1protein with each overlapping by 11 amino acids. Naive CD8� Tcells from healthy donors were stimulated with autologous DCspulsed with the peptide library and restimulated twice. Twenty-fourhours after the third stimulation, CD137� T cells were enrichedusing magnetic beads as described. One week later, enriched lineswere tested for recognition of the peptide pool, with positiveresponses then resolved by subsequent sequential stimulations of

10:1 3:1 1:1 1:30

5

10

15

20

25

30

35

E:T

% s

pe

cif

ic l

ys

is10:1 3:1 1:1 1:3

0

5

10

15

20

25

30

35

E:T

% s

pe

cif

ic l

ys

is10:1 3:1 1:1 1:3

0

5

10

15

20

25

30

35

E:T

% s

pe

cif

ic l

ys

is10:1 3:1 1:1 1:3

0

5

10

15

20

25

30

35

E:T

% s

pe

cif

ic l

ys

is

CD137

MHC-multimer CD137 MHC-multimer

CD

8

8% 86% 87%

1.2% 57%

i ii iii

iv v

A

B

Figure 5. CD137 enrichment includes tumor-lytic T cells. (A) A Melan-A–specificT-cell line was generated by 2 rounds of stimulation (i). The cells were restimulatedwith peptide-pulsed autologous PBMCs, and 24 hours later CD137� cells werepositively selected using antibody/bead complexes (ii). Selected T cells were culturedfor additional 8 days until pMHC multimer staining was performed (iii). This enrichedline was then coincubated for 24 hours (ratio 1:1) with the HLA-A2� Melan-A�

melanoma cell line A325 (iv) or with the HLA-A2� Melan-A� melanoma cell lineMel526 (v), before staining for CD137. (B) The same enriched line (filled symbols)and the CD137� flowthrough (open symbols) were also tested in a 4-hour chromiumrelease assay using the melanoma cell lines as targets (circles indicate targets;Melan-A�, Mel526 melanoma cells; and triangles, Melan A� A325 melanoma cells).

CD137 IDENTIFIES CD8 T CELLS RESPONDING TO ANTIGEN 207BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

aliquots of the responding cells deconvoluting the peptide pool(Figure 6A). Initial experiments using either no enrichment methodor the IFN�-secretion assay failed to enrich or detect any positiveresponses to the peptide pool (data not shown). By contrast, in 3 of3 experiments using naive CD8 T cells from 2 different donorsenriched for CD137� cells, 5 immunogenic epitopes were identi-fied (data not shown). One of these epitopes has been characterizedin more detail by deconvoluting the peptide pool representingWT1, and found to be contained in 2 overlapping pentadecamers,denoted as 71 and 72 (Figure 6B). To identify the HLA restrictionelement, 23 Epstein-Barr virus (EBV)–transformed, partially HLA-matched B-cell lines were pulsed with either peptide 71 or peptide72, and used as APCs. HLA-Cw0701 was identified as therestriction element, with HLA-Cw0702 also presenting the epitope,as 15 of 16 B-cell lines expressing HLA-Cw07 presented thepeptide as measured by intracellular IFN� staining, whereas 7/7HLA-Cw07–negative B-cell lines did not present the peptide(Table 3). In a T-cell line, generated from a second donor, whoshared only the Cw07-allele with the original donor, theminimal essential amino acid sequence was then determined.The WT1(286-294) (YRIHTHGVF) 9-mer induced a strong IFN�response, whereas loss of tyrosine in position 1 or the phenylal-anine in position 9 described as an anchor in the binding motiffor HLA-Cw0728 abrogated recognition (Figure 6C; Table 4).

Discussion

The biologic role of CD137 to provide a survival signal toactivated T cells has been demonstrated in vivo and in vitro.9,29

Here we describe a novel use of CD137 based on the kinetics ofexpression after antigen recognition to identify and expandantigen-specific CD8� cells. This new application has severaladvantages over alternative techniques for evaluating CD8�

T-cell responses: (1) it defines a broad and, based on pMHCmultimer binding, nearly complete repertoire of the antigen-specific CD8� T cells in a heterogeneous population, whichcannot be achieved by analyzing the production of a singlecytokine alone; (2) it allows the detection, isolation, andsubsequently the rapid expansion of antigen-specific CD8� cellseven if present initially at a very low precursor frequency; (3) itcan be performed independent of knowledge of the immuno-genic epitopes and MHC-restriction elements; (4) it correlateswell with already established assays such as CFC; and (5) itrequires a simple staining procedure of a surface antigen that hasa relatively long window of up-regulated expression afterstimulation.

Exploiting the characteristics of CD137 expression providesa broader picture of the CD8� T-cell population responding to

0 10 20 30 40 50

71

72

76

% IFN +

pe

pti

de

nu

mb

er

0 5 10 15 20 25 30

noneWT1-complete

123456789

10

% IFN +

pe

pti

de

po

ol

0 10 20 30 40 50

3 4 713 14 16

19 2229 3038 39

41 44 4550 51 54

58 60 61 64 6771 72 76

82 88 90 9396 100 102

% IFN +

pe

pti

de

nu

mb

er

Primary stimulation: peptide-pulsed (110

peptides), mature DC + naïve CD8+ T cells

1st restimulation with peptide-pulsed PBMC

CD137 enrichment

2nd restimulation with peptide-pulsed PBMC

Selection of lines responsive to the

complete peptide pool (110 peptides)

Testing of smaller peptide pools

Testing of single peptides

A

i

ii

iii

QYRIHTHGV285-292

YRIHTHGVF286-293

RIHTHGVFR287-294

CD

8

IFN

no peptide

0.2%

0.1%

88%

0.1%

B C

g

g

g

g

Figure 6. Implementing CD137 enrichment for epitope discovery. (A) Experimental outline: After 3 rounds of stimulation with a pool of 110 overlapping 15-mers, CD137�

cells were enriched and further expanded. The immunogenic peptides were subsequently identified by step-wise testing of smaller peptide pools. (Bi) An enriched T-cell linewas tested for IFN� production in response to autologous EBV-transformed B lymphoblastoid cell lines (B-LCLs) pulsed with pools of peptides (10 or 11 peptides/pool). (ii)Peptides derived from pools 3, 4 (clearly positive), and 7 (potentially positive) were tested in smaller pools containing 2 to 5 peptides, in which the peptides were rearranged byselecting partially overlapping peptide sequences. (iii) Identification of peptides 71 and 72 as the stimulating peptides. (C) Fine mapping of the minimal essential amino acidsequence was performed using a T-cell line generated from a second HLA-Cw07� donor after 3 stimulations with the immunogenic peptide. A panel of peptides wassynthesized and tested for recognition of the minimal essential amino acid (see also Table 4) using peptide-pulsed, autologous B-LCLs as stimulators.

208 WOLFL et al BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

antigen than the analysis of a single cytokine alone, asdemonstrated using naive T cells. Based on these data andpreliminary data using polarizing priming conditions, such aspriming in the presence of exogenous IL-4 (not shown), CD137appears to serve as a uniform marker for different T-cell subsetssuch as central memory cells or CD8� T cells primed underTh2/Tc2 conditions. Screening for responsive CD137� T-cellpopulations, followed by analysis of production of differentcytokines should enhance the detection, monitoring, and selec-tion of diverse responding CD8� T-cell subpopulations, andcould facilitate vaccine development requiring quantitative andqualitative characterization of the CD8� T-cell response.6,30

Our own preliminary data, as well as the work done by Wehleret al studying depletion of alloreactive T cells,14 indicate thatCD137 is also specifically up-regulated on CD4� T cells. Howeverfurther work is required to define specificity and kinetics of CD137up-regulation on CD4� cells and its comparison with selectionmethods based on expression of CD154.31,32

The ability to detect and expand ex vivo high-avidity CD8�

T cells specific for candidate tumor-associated antigens isessential not only for pursuing adoptive T-cell immunotherapybut also for validating the immunogenicity of candidate anti-gens. We have recently demonstrated the feasibility of generat-ing in vitro T-cell clones from a naive repertoire for the knownHLA-A*0201–restricted WT1(126–134) epitope.18 By using CD137expression as a means to enrich responding cells, we now extendthese findings and demonstrate that specific CD8� T cells can begenerated from a naive T-cell repertoire against previouslyundefined epitopes. By permitting early enrichment after stimu-

lation of naive cells, the likelihood of detecting and expandingclones is increased and the culture time and number ofrestimulations necessary are reduced. As each round of restimu-lation drives the cells further toward terminal differentiation or“exhaustion,”33 methods to clone reactive cells after fewerdivisions can have a direct impact on the functional quality ofthe resulting T cells. In fact, preliminary data suggest a fractionof CD8� T-cell clones generated after only 2 rounds ofstimulation have a functional and phenotypic profile resemblingearly/intermediate effector memory cells34 with maintainedCD28 expression and IL-2-production (data not shown).

The identification of novel T-cell epitopes in relativelyweakly immunogenic proteins often requires screening of a vastnumber of potential epitopes for each candidate protein. Byusing APCs pulsed with a pool of peptides encompassing thewhole WT1 protein and enriching and testing the respondingCD137� cells, it has proved possible to generate and character-ize novel T-cell responses against this weakly immunogenicself-antigen. The use of such a peptide library for the analysisand expansion of memory T cells against CMV35–37- or HIV38-derived proteins has been described recently, and offers theadvantage that it does not require an a priori selection ofcandidate peptides. For example, in silico analysis, predictingpeptides from within the protein with favorable binding charac-teristics to a selected HLA-molecule, has often been used tonarrow the number of candidate peptides,39 but prediction ofbinding scores, especially for less common HLA alleles, issometimes unreliable or not available. In fact, the newlyidentified epitope WT1(286-294) models at only 8.2% of themaximal possible score, ranking as the 25th best peptide forHLA-Cw0702 from the complete amino acid sequence of WT1as computed using the BIMAS algorithm40 (the other commonlyused algorithm, SYFPEITHI,41 does not provide calculations forthis allele) and would likely have been missed if peptides hadbeen selected on the basis of in silico prediction as demonstratedfor other antigens.42

Adoptive immunotherapy with antigen-specific CD8� T cells hasbeen shown to be effective in the treatment of human tumors,43 but islimited by the time and difficulty of isolating and expanding functionalcells.24 Implementing novel technologies that facilitate obtaining a

Table 4. Fine mapping of WT1(286-293) as the minimal essential aminoacid sequence using peptide-pulsed, autologous B-LCLs as APCs

Peptideposition Sequence

Recognition(5% threshold)

277-291 TTPILCGAQYRIHTH �

281-295 LCGAQYRIHTHGVFR �

285-299 QYRIHTHGVFRGIQD �

286-300 YRIHTHGVFRGIQDV �

289-303 HTHGVFRGIQDVRRV �

284-294 AQYRIHTHGVFR �

284-293 AQYRIHTHGVF �

284-292 AQYRIHTHGV �

285-294 QYRIHTHGVFR �

285-293 QYRIHTHGVF �

285-292 QYRIHTHGV �

286-294 YRIHTHGVFR �

286-293 YRIHTHGVF �

287-294 RIHTHGVFR �

T cells generated by stimulation with the 15-mer for WT-1(281-295) were tested byCFC for recogniton of autologous B-LCLs pulsed with the indicated peptides(10 �g/mL).

Table 3. Presentation of WT1(286-293) by peptide-pulsed B-LCLs withpartially matching HLA type

No. Matching allele Nonmatching alleles

Original donor: A1/A24/B8/B40/Cw03/Cw07*

B-cell-lines not presenting the epitope (IFN�� T cells <5%)

1 A1 A30/B44/B57/Cw5/Cw6

2 A1 A11/B53/B57/Cw4/Cw6

3 A24 A32/B35/B44/Cw4/Cw4

4 A24/Cw3 A68/B35/B44/Cw4

5 A24/Cw3 A3/B15/B35/Cw4

6 — A30/A32/B27/B50/Cw1/Cw4

7 A24/B40 A32/B35/Cw2/Cw4

8 B40/Cw7/Cw7*† A203/A20/B38

B-cell-lines presenting the epitope (IFN�� T cells >5%)

9 A1/B8/Cw7/Cw7* A2/B18

10 A1/B8/Cw7/Cw7 A29/A44

11 A1/B8/Cw7/Cw7 A2/B7

12 A1/B8/Cw7/Cw7 A3/B49

13 A1/B8/Cw7/Cw7* A3/B49

14‡ A1/B8/Cw3/Cw7 A2/B55

15 A24/B40/Cw1/ Cw7* A32/B35

16 B40/Cw3/Cw7† A3/A3/B39

17 B40/Cw7/Cw7 A203/A20/B38

18 Cw7/Cw7*† A2/A3/B7/B18

19 Cw72 A3/A3/B7/B41/Cw17

20 Cw7* A2/A3/B18/B44/Cw12

21 Cw7† A3/A29/B7/B44/Cw16

22‡ Cw7/Cw7 A2/A2/B7/B7

23‡ Cw7* A2/A23/B27/B49/C1

T cells specific for WT-1(286-293) were tested for recogniton of peptide-pulsedB-LCLs with partially matching HLA-type and subsequent INF� production (CFC).

*HLA-Cw0701.†HLA-Cw0702.‡Presentation only at 25 �g/mL peptide.

CD137 IDENTIFIES CD8 T CELLS RESPONDING TO ANTIGEN 209BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

larger repertoire of specific T cells, accomplishing this in a faster morecost-efficient way, and generating cells that can exhibit broader func-tional profiles will facilitate broader and more effective application ofthis strategy. The CD137-enrichment method should prove useful forhelping to achieve these goals.

Acknowledgments

This work was supported by grants from the Leukemia &Lymphoma Society (LLS 7040–03) and the National Institute ofHealth (P01 CA18029, R37 CA33084). M.W. and J.K. are fellowsof the Deutsche Krebshilfe (Germany).

The authors gratefully acknowledge expert technical help fromNatalie Duerkopp and Audrey Mollerup.

Authorship

Contribution: M.W. designed, performed, and analyzed experi-ments; J.K., H.N., W.Y.H., and P.D.G. designed and analyzedexperiments; T.J.M. and M.B. provided vital reagents; and M.W.and P.D.G. wrote the paper. All authors edited and approved thewritten paper.

Conflict-of-interest disclosure: The authors declare no compet-ing financial interests.

Correspondence: Matthias Wolfl, Pediatric Stem Cell TransplantProgram, University Children’s Hospital Wuerzburg, Josef-Schneider-Strasse 2, Wuerzburg D-97080, Germany; e-mail: woelfl_m@klinik.uni-wuerzburg.de or mwolfl@arcor.de; Philip D. Greenberg, University ofWashington, BB1325 Health Sciences Building, Box 356527, Seattle,WA 98195; e-mail: pgreen@u.washington.edu.

References

1. Altman JD, Moss PA, Goulder PJ, et al. Pheno-typic analysis of antigen-specific T lymphocytes.Science. 1996;274:94-96.

2. Manz R, Assenmacher M, Pfluger E, Miltenyi S,Radbruch A. Analysis and sorting of live cells ac-cording to secreted molecules, relocated to a cell-surface affinity matrix. Proc Natl Acad Sci U S A.1995;92:1921-1925.

3. Betts MR, Brenchley JM, Price DA, et al. Sensitiveand viable identification of antigen-specific CD8� Tcells by a flow cytometric assay for degranulation.J Immunol Methods. 2003;281:65-78.

4. Mosmann TR, Li L, Sad S. Functions of CD8 T-cell subsets secreting different cytokine patterns.Semin Immunol. 1997;9:87-92.

5. Sallusto F, Geginat J, Lanzavecchia A. Centralmemory and effector memory T cell subsets:function, generation, and maintenance. Annu RevImmunol. 2004;22:745-763.

6. Kern F, LiPira G, Gratama JW, Manca F, RoedererM. Measuring Ag-specific immune responses: under-standing immunopathogenesis and improving diag-nostics in infectious disease, autoimmunity and can-cer. Trends Immunol. 2005;26:477-484.

7. Kwon BS, Weissman SM. cDNA sequences oftwo inducible T-cell genes. Proc Natl Acad SciU S A. 1989;86:1963-1967.

8. Vinay DS, Kwon BS. Role of 4-1BB in immuneresponses. Semin Immunol. 1998;10:481-489.

9. Watts TH. TNF/TNFR family members in costimu-lation of T cell responses. Annu Rev Immunol.2005;23:23-68.

10. Cannons JL, Lau P, Ghumman B, et al. 4-1BB ligandinduces cell division, sustains survival, and enhanceseffector function of CD4 and CD8 T cells with similarefficacy. J Immunol. 2001;167:1313-1324.

11. Lee HW, Park SJ, Choi BK, Kim HH, Nam KO,Kwon BS. 4-1BB promotes the survival of CD8�T lymphocytes by increasing expression of Bcl-xLand Bfl-1. J Immunol. 2002;169:4882-4888.

12. Dawicki W, Watts TH. Expression and function of4-1BB during CD4 versus CD8 T cell responsesin vivo. Eur J Immunol. 2004;34:743-751.

13. Cho HR, Kwon B, Yagita H, et al. Blockade of 4-1BB(CD137)/4-1BB ligand interactions increases allo-graft survival. Transpl Int. 2004;17:351-361.

14. Wehler TC, Nonn M, Brandt B, et al. Targeting theactivation-induced antigen CD137 can selectively de-plete alloreactive T cells from antileukemic and antitu-mor donor T-cell lines. Blood. 2007;109:365-373.

15. Pittet MJ, Valmori D, Dunbar PR, et al. High fre-quencies of naive Melan-A/MART-1-specificCD8(�) T cells in a large proportion of humanhistocompatibility leukocyte antigen (HLA)-A2individuals. J Exp Med. 1999;190:705-715.

16. Keilholz U, Menssen HD, Gaiger A, et al. Wilms’tumour gene 1 (WT1) in human neoplasia. Leu-kemia. 2005;19:1318-1323.

17. Van Driessche A, Gao L, Stauss HJ, et al. Anti-gen-specific cellular immunotherapy of leukemia.Leukemia. 2005;19:1863-1871.

18. Ho WY, Nguyen HN, Wolfl M, Kuball J, GreenbergPD. In vitro methods for generating CD8(�) T-cellclones for immunotherapy from the naive reper-toire. J Immunol Methods. 2006;310:40-52.

19. Jung T, Schauer U, Heusser C, Neumann C, RiegerC. Detection of intracellular cytokines by flow cytom-etry. J Immunol Methods. 1993;159:197-207.

20. Pulle G, Vidric M, Watts TH. IL-15-dependent induction of4-1BB promotes antigen-independent CD8 memory T cellsurvival. J Immunol. 2006;176:2739-2748.

21. Cox AL, Mosbruger T, Mao Q, et al. Cellular im-mune selection with hepatitis C virus persistencein humans. J Exp Med. 2005;201:1741-1752.

22. van Lier RA, ten Berge IJ, Gamadia LE. HumanCD8(�) T-cell differentiation in response to vi-ruses. Nat Rev Immunol. 2003;3:931-939.

23. Pittet MJ, Zippelius A, Speiser DE, et al. Ex vivoIFN-gamma secretion by circulating CD8 T lym-phocytes: implications of a novel approach for Tcell monitoring in infectious and malignant dis-eases. J Immunol. 2001;166:7634-7640.

24. Ho WY, Blattman JN, Dossett ML, Yee C, Green-berg PD. Adoptive immunotherapy: engineering Tcell responses as biologic weapons for tumormass destruction. Cancer Cell. 2003;3:431-437.

25. Oka Y, Elisseeva OA, Tsuboi A, et al. Human cy-totoxic T-lymphocyte responses specific for pep-tides of the wild-type Wilms’ tumor gene (WT1)product. Immunogenetics. 2000;51:99-107.

26. Azuma T, Makita M, Ninomiya K, Fujita S, HaradaM, Yasukawa M. Identification of a novel WT1-derived peptide which induces human leucocyteantigen-A24-restricted anti-leukaemia cytotoxic Tlymphocytes. Br J Haematol. 2002;116:601-603.

27. Bellantuono I, Gao L, Parry S, et al. Two distinctHLA-A0201-presented epitopes of the Wilms tu-mor antigen 1 can function as targets for leuke-mia-reactive CTL. Blood. 2002;100:3835-3837.

28. Falk K, Rotzschke O, Grahovac B, et al.Allele-specificpeptide ligand motifs of HLA-C molecules. Proc NatlAcad Sci U SA. 1993;90:12005-12009.

29. Wen T, Bukczynski J, Watts TH. 4-1BB ligand-mediated costimulation of human T cells inducesCD4 and CD8 T cell expansion, cytokine produc-tion, and the development of cytolytic effectorfunction. J Immunol. 2002;168:4897-4906.

30. Wille-Reece U, Flynn BJ, Lore K, et al. Toll-likereceptor agonists influence the magnitude andquality of memory T cell responses after prime-

boost immunization in nonhuman primates. J ExpMed. 2006;203:1249-1258.

31. Chattopadhyay PK, Yu J, Roederer M. A live-cell as-say to detect antigen-specific CD4� T cells with di-verse cytokine profiles. Nat Med. 2005;11:1113-1117.

32. Frentsch M, Arbach O, Kirchhoff D, et al. Directaccess to CD4� T cells specific for defined anti-gens according to CD154 expression. Nat Med.2005;11:1118-1124.

33. Gattinoni L, Klebanoff CA, Palmer DC, et al. Ac-quisition of full effector function in vitro paradoxi-cally impairs the in vivo antitumor efficacy ofadoptively transferred CD8� T cells. J Clin In-vest. 2005;115:1616-1626.

34. Rufer N, Zippelius A, Batard P, et al. Ex vivo char-acterization of human CD8� T subsets with dis-tinct replicative history and partial effector func-tions. Blood. 2003;102:1779-1787.

35. Kern F, Surel IP, Brock C, et al. T-cell epitope map-ping by flow cytometry. Nat Med. 1998;4:975-978.

36. Kern F, Faulhaber N, Frommel C, et al. Analysisof CD8 T cell reactivity to cytomegalovirus usingprotein-spanning pools of overlapping pentade-capeptides. Eur J Immunol. 2000;30:1676-1682.

37. Trivedi D, Williams RY, O’Reilly RJ, Koehne G.Generation of CMV-specific T lymphocytes usingprotein-spanning pools of pp65-derived overlap-ping pentadecapeptides for adoptive immuno-therapy. Blood. 2005;105:2793-2801.

38. Draenert R, Verrill CL, Tang Y, et al. Persistent recog-nition of autologous virus by high-avidity CD8 T cellsin chronic, progressive human immunodeficiencyvirus type 1 infection. J Virol. 2004;78:630-641.

39. Rammensee HG, Weinschenk T, GouttefangeasC, Stevanovic S. Towards patient-specific tumorantigen selection for vaccination. Immunol Rev.2002;188:164-176.

40. Parker KC, Bednarek MA, Coligan JE. Schemefor ranking potential HLA-A2 binding peptidesbased on independent binding of individual pep-tide side-chains. J Immunol. 1994;152:163-175.

41. Rammensee H, Bachmann J, Emmerich NP, Ba-chor OA, Stevanovic S. SYFPEITHI: database forMHC ligands and peptide motifs. Immunogenet-ics. 1999;50:213-219.

42. Pelte C, Cherepnev G, Wang Y, Schoenemann C,Volk HD, Kern F. Random screening of proteinsfor HLA-A*0201-binding nine-amino acid peptidesis not sufficient for identifying CD8 T cell epitopesrecognized in the context of HLA-A*0201. J Im-munol. 2004;172:6783-6789.

43. Rosenberg SA, Dudley ME. Cancer regression inpatients with metastatic melanoma after the transferof autologous antitumor lymphocytes. Proc Natl AcadSci U S A. 2004;101(suppl 2):14639-14645.

210 WOLFL et al BLOOD, 1 JULY 2007 � VOLUME 110, NUMBER 1

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom

online March 19, 2007 originally publisheddoi:10.1182/blood-2006-11-056168

2007 110: 201-210  

Philip D. GreenbergMatthias Wolfl, Jürgen Kuball, William Y. Ho, Hieu Nguyen, Thomas J. Manley, Marie Bleakley and antigen without requiring knowledge of epitope specificities

T cells responding to+and expansion of the full repertoire of CD8Activation-induced expression of CD137 permits detection, isolation, 

http://www.bloodjournal.org/content/110/1/201.full.htmlUpdated information and services can be found at:

(5403 articles)Immunobiology    Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society

For personal use only.on August 12, 2016. by guest www.bloodjournal.orgFrom