Amino-Terminal Extended Peptide Single-Chain Trimers Are Potent Synthetic Agonists for Memory Human CD8+ T Cells

This information is current as Beatriz M. Carreno, Michelle Becker-Hapak, Megan Chan, of September 26, 2021. Wen-Rong Lie, Xiaoli Wang, Ted H. Hansen and Gerald P. Linette J Immunol 2012; 188:5839-5849; Prepublished online 9 May 2012;

doi: 10.4049/jimmunol.1103647 Downloaded from http://www.jimmunol.org/content/188/12/5839

Supplementary http://www.jimmunol.org/content/suppl/2012/05/09/jimmunol.110364 Material 7.DC1 http://www.jimmunol.org/ References This article cites 43 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/188/12/5839.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Amino-Terminal Extended Peptide Single-Chain Trimers Are Potent Synthetic Agonists for Memory Human CD8+ T Cells

Beatriz M. Carreno,* Michelle Becker-Hapak,* Megan Chan,* Wen-Rong Lie,† Xiaoli Wang,‡ Ted H. Hansen,‡ and Gerald P. Linette*

Upon Ag exposure, most memory T cells undergo restimulation-induced cell death. In this article, we describe a novel synthetic agonist, an N-terminal extended decamer peptide expressed as a single-chain trimer, the amino-terminal extended peptide MHC class I single-chain trimer (AT-SCT), which preferentially promotes the growth of memory human CD8+ T cells with minimal restimulation-induced cell death. Using CMV pp65 and melanoma gp100 Ags, we observe the in vitro numerical expansion of a clonally diverse polyfunctional population of Ag-specific CD8+ T cells from healthy individuals and vaccinated melanoma patients, respectively. Memory CD8+ T cells stimulated with AT-SCT presented on MHC class I/II-null cells show reduced production, slower kinetics of TCR downregulation, and decreased cell death compared with native nonamer MHC Downloaded from class I single-chain trimer (SCT)-activated T cells. However, both ERK phosphorylation and cell cycle kinetics are identical in AT- SCT– and SCT-activated T cells. Probing of SCT and AT-SCT peptide–MHC complexes using fluorochrome-conjugated TCR multimers suggests that nonamer- and decamer-linked peptides may be anchored differently to the HLA-A2 peptide-binding groove. Our findings demonstrate that modified peptide–MHC structures, such as AT-SCT, can be engineered as agonists to promote the growth and expansion of memory human CD8+ T cells. The Journal of Immunology, 2012, 188: 5839–5849. http://www.jimmunol.org/

he specificity of T cell activation is largely dictated by for activation (13, 14) and are more sensitive to cell death com- the interaction of the TCR with peptide–MHC (p-MHC) pared with naive resting T cells upon Ag exposure (15–17). T (1). TCR–p-MHC interactions can lead to a spectrum of MHC class I single-chain trimers (SCT) are fusion proteins, responses in mature T cells ranging from full activation, resulting which consist of an antigenic peptide-flexible linker–b2-micro- in clonal expansion (or cell death), to a state of anergy, resulting in globulin (b2m)-flexible linker-class I H chain polypeptide (18). unresponsiveness (2–4). These distinct outcomes have been at- The essential properties of SCT are multifold and include pres- tributed, in part, to differences in binding affinity or off-rates of ervation of the native structure of peptide/b2m/MHC class I, as the interaction between TCR and p-MHC (5–8). Single amino acid detected by CD8+ T cells and conformational-dependent mAbs, by guest on September 26, 2021 substitutions can create structural variants of antigenic peptide such as 25-D1.16 (19). Crystallographic data with the SIINFEKL/ b ligands, referred to as altered peptide ligands (APL), which allow b2m/K SCT confirm structural equivalence with free peptide bound b b manipulation of the strength of the TCR–p-MHC interaction. APL to K (20) and, importantly, SCT encoding SIINFEKL/b2m/K can induce a complete range of T cell functions (agonist) or a subset stimulates IL-2 secretion and vigorous proliferation of primed of T cell functions (partial agonist) (9, 10); in a few instances, it can CD8+ OT-1 transgenic T cells (21). Additionally, several indepen- be more potent (superagonist) than the native ligand (11, 12). It is dent studies demonstrated that, as DNA , SCT can elicit ro- also well established that memory T cells display a lower threshold bust CD8+ T cell responses to well-defined Ags in murine models (19). Recently, we reported on the biochemical stability of HLA *Division of Oncology, Department of Medicine, Washington University School class I-containing SCT and their serological recognition by of Medicine, St. Louis, MO 63110; †EMD Millipore, Billerica, MA 01822; and conformational-dependent Abs, as well as demonstrated their ‡ Department of Pathology and Immunology, Washington University School of Med- recognition by Ag-specific T cells (22). In the current study, we icine, St. Louis, MO 63110 examine whether SCT, expressed in the MHC class I/II-null cell Received for publication December 15, 2011. Accepted for publication April 5, 2012. line K562, may represent a p-MHC entity for ex vivo stimulation This work was supported in part by a grant from Washington University Bear Cub + Fund (to G.P.L.), a grant from the Barnes-Jewish Hospital Foundation (to G.P.L.), and of memory (Ag-primed) human CD8 T cells. From a series of National Institutes of Health Grant AI055849 (to T.H.H.). The Siteman Cancer Cen- SCT constructs generated, a new p-MHC entity, the amino-ter- ter High-Speed Cell Sorter Core Facility is supported in part by National Cancer minal extended peptide MHC class I single-chain trimer (AT-SCT), Institute Cancer Center Support Grant P30 CA91842. was identified as a potent inducer of human CD8+ T cell expansion Address correspondence and reprint requests to Dr. Beatriz M. Carreno, Division of Oncology, Washington University School of Medicine, 660 South Euclid Avenue, and growth. AT-SCT for two well-characterized Ags are described Box 8007, St. Louis, MO 63110. E-mail address: [email protected] in detail: the CMV pp65 antigenic peptide (NLVPMVATV) and the The online version of this article contains supplemental material. melanoma gp100 G209-2M antigenic peptide (IMDQVPFSV). AT- Abbreviations used in this article: 7AAD, 7-aminoactinomycin D; APL, altered SCT delivers a TCR signal distinct from conventional native SCT, peptide ligand; AT-SCT, amino-terminal extended peptide MHC class I single- resulting in .400-fold expansion of Ag-specific human CD8+ chain trimer; AT-SCT dt, disulfide trap amino-terminal extended peptide MHC class T cells from healthy donors, as well as melanoma patients. I single-chain trimer; DC, dendritic cell; FSC, forward light scatter; b2m, b2-micro- globulin; mDC, mature dendritic cell; p-MHC, peptide–MHC; RICD, restimulation- induced cell death; RPMI-5, RPMI 1640 supplemented with 5% pooled human sera; SCD, single-chain dimer; SCT, MHC class I single-chain trimer; SCT dt, disulfide Materials and Methods trap MHC class I single-chain trimer; SSC, side scatter; Stemline-5, Stemline media Generation of SCT-expressing K562 transfectants supplemented with 5% pooled human sera. HLA-A*0201 SCT were generated, as previously described (22). For AT- Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 SCT generation, synthetic DNA oligonucleotides (Integrated DNA Tech- www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103647 5840 SYNTHETIC AGONISTS FOR CD8+ T CELLS nologies, Coralville, IA) encoding 10-mer antigenic peptides extended at of cytolytic granule content, cells were fix/permeabilized with Cytofix/ the N terminus were ligated into the SCT vectors at restriction sites spe- Cytoperm (BD), according to the manufacturer’s instructions, and stain- cifically designated for peptide sequence shuffling. Multiple amino acid ed with Abs to perforin, granzyme A (BioLegend), and granzyme B residues were introduced at the N terminus of optimal 9-mer peptide to (Invitrogen). Polyfunctionality of expanded Ag-specific T cells was de- generate AT-SCT. For NLV (CMV pp65), constructs were made containing termined as follows: 14 d after SCT or AT-SCT stimulation, T cells were the naturally occurring amino acid residue (ANLV, underline identifies restimulated with SCT at 1:1 ratio in RPMI 1640 supplemented with 5% amino acid residue at N terminus of peptide in AT-SCT) or amino acids pooled human sera (RPMI-5) and brefeldin A (10 mg/ml; eBioscience). As with bulky nonpolar (YNLV) and uncharged polar (NNLV) R groups, as controls, cells were cultured in media in the absence of stimulation. Six- well as nonbulky uncharged polar (GNLV) R group. The ANLV construct teen hours after stimulation, cells were harvested, washed, fix/perme- was poorly expressed on the cell surface and was not studied further; all abilized with Cytofix/Cytoperm, and stained with pretitered Abs specific other AT-SCT constructs were expressed on the cell surface at levels for IFN-g (BioLegend), TNF-a, IL-2 (eBioscience), and MIP-1b (BD). similar to the SCT. For IMD (G209-2M), constructs containing the natu- Cells were then washed, fixed with 2% paraformaldehyde, and analyzed rally occurring amino acid residue (TIMD) were made; for comparison by polychromatic flow cytometry. Functional complexity of Ag-specific with the CMV system, an AT-SCT construct expressing the amino acid responses was assessed as described (25). The proportion of CD8+ with bulky uncharged polar (NIMD) R group was made. A construct T cells contributing to each response pattern was determined, grouped, and containing human b2m and HLA-A*0201 with no covalently linked pep- color-coded according to the number of positive functions. Slices in the tide, referred to as a single-chain dimer (SCD), was also generated. pie charts represent the proportion of responding CD8+ T cells that Generation of disulfide trap SCT (SCT dt) molecules was performed, as have produced all four or any combination of three, two, or previously described (22). All constructs were verified by DNA sequenc- one cytokines. For analysis of cytokines secreted by T cells upon SCT ing. K562 cells (American Type Culture Collection, Manassas, VA) were activation, cultures were activated, as described above, in the absence transfected with constructs using Nucleofector kit 5 (Lonza, Basel, Swit- of brefeldin A, and supernatants were collected 24 h after activation zerland), and stable cell lines were selected and maintained in 0.5 mg/ml and characterized using a Milliplex cytokine kit (Millipore, Billerica, MA),

Geneticin (Life Technologies, Grand Island, NY). Transfectants were per the manufacturer’s instructions. For characterization of TCR Vb usage Downloaded from assessed for HLA-A2 expression using the mAb BB7.2 (Serotec, by Ag-specific CD8+ T cells after AT-SCT expansion, T cells were purified Kidlington, U.K.). Expression of NLV (CMV pp65)/A*0201 and IMD using a cell sorter (98% tetramer+/CD8+), and total RNA was extracted (G209-2M)/A*0201 peptide complexes on the surface of K562 trans- (RNeasy, QIAGEN, Valencia, CA), followed by a one-step cDNA syn- fectants was assessed using PE-conjugated TCR multimers (Altor Bio- thesis using nested primers specific for each of the 22 TCR Vb family science, Miramar, FL) (23), as per the manufacturer’s instructions. members (BioMed Immunotech, Tampa, FL). TCR clonality was deter- mined using a 4% high-resolution gel, as per the manufacturer’s instruc- Human donors tions (BioMed Immunotech). was performed using a BD Bioscience (San Jose, CA) FACScan flow cytometer (upgraded by Cytek http://www.jimmunol.org/ Leukapheresis was performed to obtain peripheral blood leukocytes from Development [Freemont, CA] to three lasers and 10 fluorescent parameters normal donors through the Barnes-Jewish Hospital blood bank. Donors plus two scatters). Data were acquired and analyzed using FlowJo software were tested for CMV serology and HLA class I typed. Leukapheresis was (Ashland, OR). also performed from melanoma patients enrolled in a phase I clinical trial (NCT00683670; Dendritic Cell Vaccination for Advanced Melanoma). Apoptosis and CFSE Written informed consent was obtained from all healthy donors and melanoma patients prior to any study procedures, and both studies were For apoptosis determination, CMV pp65 (NLV)-specific T cells (90% approved by the Siteman Cancer Center Protocol Review and Monitoring tetramer+) were stimulated with irradiated SCT or AT-SCT at a ratio of 1:1 Committee and the Washington University Human Research Protocol in RPMI-5 for the indicated time, harvested, and stained with FITC Organization. conjugated-annexin using binding buffer, followed by addition of 7-ami- by guest on September 26, 2021 noactinomycin D (7AAD) and analysis by flow cytometry. A total of 2.5 3 Generation and expansion of Ag-specific T cells 104 events was collected in the side scatter (SSC)/forward light scatter (FSC) gate. For CFSE analysis, cells were labeled using the CellTrace Dendritic cells (DC) were generated from monocyte cultures in GM-CSF CFSE kit (Invitrogen), as per the manufacturer’s instructions, and stimu- (100 ng/ml) and IL-4 (20 ng/ml) and matured using CD40L and IFN-g (100 + lated as described above. At the indicated time, cells were harvested and U/ml), as previously described (24). CD8 T cells were isolated from analyzed by flow cytometry. PBMCs using a CD82 negative-selection kit (Miltenyi Biotec, Auburn, + 3 CA); purity was 85–90% (data not shown). Purified CD8 T cells (0.5 Determination of ERK phosphorylation 105/ml) were cultured at a 20:1 ratio with irradiated (2500 rad) autologous mature DC (mDC) pulsed with peptide (2 mg/106 DC/ml) in 48-well trays To assess ERK phosphorylation upon SCT or AT-SCT stimulation, NLV- in Stemline media (S1694; Sigma-Aldrich, St. Louis, MO) supplemented specific T cells (94% tetramer+) were allowed to equilibrate for 1 h at with 5% pooled human sera (Stemline-5). Peptides used were CMV pp65 37˚C in RPMI-5 before stimulators were added (1:10 T/SCT ratio); they (NLVPMVATV) and melanoma human gp100-derived peptide G209-2M were spun down to pellet cells and incubated for the indicated time at (IMDQVPFSV). Human IL-2 (10–50 U/ml; Chiron, Emeryville, CA) was 37˚C, followed by fixation in 4% paraformaldehyde, permeabilization added every 2 d, starting 48 h after culture initiation. Fourteen days after with 90% methanol, and staining with anti-phospho ERK Ab (clone 20A; DC stimulation, T cell cultures were harvested, characterized for Ag- BD Bioscience). Unstimulated and PMA/ionomycin-stimulated T cells specific frequencies using HLA/peptide tetramers (see below), and were used as negative and positive controls, respectively. Cells were restimulated with irradiated (10,000 rad) SCT- or AT-SCT–expressing analyzed by flow cytometry, as described above. K562 cells at a 1:1 ratio. Cultures were initiated in either six-well plates (106 each T and SCT or AT-SCT) or T25 flask (5 3 106 each) using [51Cr]-release and peptide-competition assays Stemline-5. Twenty-four hours after stimulation, cultures were supple- mented with IL-2 (500 U/ml), and viable cell counts were performed daily. Protocols for chromium-release assays to measure specific lysis were de- 3 5 scribed (24, 26). For direct recognition by NLV-specific CTL, AT-SCT–- Cell concentrations were maintained at 5 10 /ml throughout the culture 51 period. For large-scale expansion, T cells were cultured in gas-permeable and SCT-expressing K562 cells were labeled with 25 mCi [ Cr] for 1 h, washed, and used in a standard 4-h assay. For recognition of CMV-infected Lifecell bags (Nexell Therapeutics, Emeryville, CA). On days 10–14 of + secondary stimulation, the percentage of tetramer+ cells and the number of cells, T98G (HLA-A*0201 ) glioblastoma cells were infected using su- viable cells were used to determine tetramer yields and tetramer folds, as pernatants derived from human foreskin fibroblast cells expressing a bac- . terial artificial chromosome containing full-length human CMV strain described previously (24) and in the legend of Fig. 1. T cells from 15 51 healthy donors and 4 melanoma patients were analyzed during the course AD169 for the indicated time and labeled with 25 mCi [ Cr] for 1 h prior of this study. to a standard 4-h assay (27). For direct recognition by G209-2M–specific CTL, melanoma cell lines DM6 (HLA-A2+/gp100+) and A375 (HLA- + 2 51 Tetramer assay, phenotypic analysis, and cytokine production A2 /gp100 ) were labeled with 25 mCi [ Cr] in the presence or absence of G209-2M peptide (5 mg/106 target) for 1 h, washed, and used in a Tetramer analysis was done as previously described (24). Flow cytom- standard 4-h assay. For peptide-competition assays (22), SCT, AT-SCT, and etry analysis was done using the following fluorochrome-conjugated HLA-A2 SCD were incubated for 2 h with increasing concentrations of anti-human mAbs: CD25, CD27, CD45RO (Invitrogen, Carlsbad, CA), G209-2M peptide, followed by labeling with 25 mCi [51Cr] for 1 h. After CD28, CCR7 (BD, Franklin Lakes, NJ), CD57, CD69, CCR5 (BioLegend, incubation, targets were washed twice and tested against a G209-2M– San Diego, CA), and PD-1 (eBioscience, San Diego, CA). For assessment specific CTL line (26) in a standard 4-h chromium-release assay. The Journal of Immunology 5841

Results specific (as determined using NLV/A*0201 tetramer) CD8+ Tcell Amino-terminal extended SCT promotes expansion of memory frequencies of 1.11%. After a 14-d stimulation with peptide- (Ag-primed) human CD8+ T cells pulsed autologous mDC, the frequency of NLV-specific CD8+ . To evaluate ex vivo proliferative responses of memory (Ag- T cells increased significantly to 40% (Fig. 1A, middle panel); + primed) CD8+ T cells upon restimulation, we chose to study the this memory CD8 T cell population (Supplemental Fig. 1) (29) HLA-A*0201–restricted/CMV pp65-specific NLV 495–503 pep- was used for restimulation assays, using SCT, as depicted in Fig. tide system in healthy CMV-seropositive donors (28). To generate 1A and as described in this study. memory CD8+ T cells, purified resting peripheral blood CD8+ To examine whether SCT may represent a p-MHC entity for T cells were stimulated in vitro with NLV peptide-pulsed autol- ex vivo stimulation of memory human CD8+ T cells, we elected ogous mDC for 14 d (24). As shown in Fig. 1A (left panel), a to express the SCT constructs (Fig. 1B) in K562 cells, a human representative healthy individual had, at baseline, CMV pp65- leukemia cell line void of HLA class I/II molecules (Supplemental Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 1. AT-SCT promotes the growth of Ag-specific T cells. CD8+ T cells purified from PBMC of seropositive CMV individuals were stimulated with CD40L/IFN-g–activated NLV-peptide–pulsed DC (mDC). Fourteen days after the mDC stimulation, T cells were restimulated with SCT or AT-SCT expressing K562 cells at a 1:1 ratio, and cultures were maintained as described in Materials and Methods. Results were obtained using SCT (NLV) and AT- SCT (NNLV) containing an asparagine as amino terminal extended residue (Materials and Methods). (A) T cell culture scheme and frequency of NLV- specific (NLV/A*0201 tetramer+) T cells in purified CD8 (left panel) and after mDC stimulation (day 14, middle panel) for a representative donor. Memory T cells were then restimulated with either SCT- or AT-SCT–expressing K562 cells for 14 d and reanalyzed on day 28 by NLV/A*0201 tetramer and anti- CD8 mAb staining. (B) Structural features of HLA-A2 single-chain constructs. A schematic representation of SCD, SCT, and AT-SCT is shown. SCT and AT-SCT containing disulfide traps to anchored antigenic peptide are depicted as SCT dt and AT-SCT dt, respectively. Peptide amino acid sequences are shown; b2m and HLA-A2 class I H chain are represented by boxes. Linkers are represented by horizontal lines, and the cysteine bond in disulfide trap molecules is represented by brackets. SCD lack a peptide sequence (22). AT-SCT contains decamer peptide sequences extended at amino termini as depicted, X represents one of three amino acids: (N) asparagine (this figure), (G) glycine, or (Y) tyrosine (Supplemental Fig. 2). (C) Growth curves of memory T cell cultures stimulated with SCT and AT-SCT from donor shown in (A). Total viable T cell numbers (left y-axis) and Ag-specific T cell yields (right y-axis) at day 14 are indicated. Each point represents single cell counts obtained at the indicated time point from SCT- or AT-SCT–stimulated cultures. (D–F) Summary of NLV-specific T cell expansion of AT-SCT– and SCT-stimulated cultures obtained from eight healthy CMV-seropositive individuals. (D) Ag-specific yields were calculated as the percentage of NLV/A*0201 tetramer+ CD8+ T cells 3 total CD8+ T cell numbers at day 14 (p = 0.0371). (E) Ag-specific frequencies were determined by flow cytometry using anti-CD8 mAb and NLV/A*0201 tetramer+ (p = 0.0054). (F) Ag- specific T cell folds represent [NLV/A*0201 tetramer+ T cell yields/frequency of NLV/A*0201 tetramer+ CD8+ T cells] 3 total CD8+ T cells at initiation of secondary stimulation. p = 0.0015, Student t test; p values , 0.05 are significant. 5842 SYNTHETIC AGONISTS FOR CD8+ T CELLS

Fig. 1). Previous work demonstrated the usefulness of K562, as an peptide sequence (Fig. 1B). We tested a variety of constructs APC, to expand human CD4+ and CD8+ T cells in vitro (30–33). (data not shown) and found that N-terminal extended decamer The parental K562 cells used for our experiments were selected peptides in the SCT format, referred to as AT-SCT, were most for expression of CD54, CD58, CD80, and CD86 (Supplemental stimulatory for NLV-specific T cells. Restimulation of memory Fig. 1). NLV-specific memory CD8+ T cells were restimulated CD8+ T cells with soluble peptides (nonamer or decamer) did not with SCT-expressing K562 cells at a 1:1 ratio and cultured for 12– lead to the numerical expansion of Ag-specific T cells obtained 14 d in media supplemented with IL-2. No other cells or cytokines with AT-SCT (Supplemental Fig. 2). Additional experiments were added to the cultures. were performed comparing memory CD8+ Tcellgrowthupon Upon restimulation, nonamer (NLVPMVATV) peptide ex- restimulation with SCT and AT-SCT; representative results using pressed as SCT failed to significantly increase the Ag-specific NLV-specific T cells are shown in Fig. 1A and 1C. In this ex- frequency (Fig. 1A, top right panel), and it modestly increased ample, 0.5 3 106 (0.22 3 106 NLV-specific) CD8+ memory Ag-specific T cell yields (Fig. 1C, see below). We interpreted the T cells were stimulated with SCT- or AT-SCT–expressing cells, previous finding as evidence that SCT, expressed in K562 cells, and cultures were continued for 14 d, as described in Materials provide an activation signal leading to restimulation-induced cell and Methods. Live cell growth kinetics were determined by death (RICD), reminiscent of the observation that Ag stimulation manual counting using trypan blue exclusion; at the termination can promote clonal deletion of mature memory T cells (2, 16). of culture, surviving cells were stained with NLV/A*0201 tetra- We postulated that modification of the peptide to create a partial mer and anti-CD8 mAb to assess the frequency of Ag-specific agonist would lead to reduced RICD (34). To test this hypothesis, CD8+ T cells by flow cytometry (Fig. 1A) and calculate Ag- a series of SCT constructs was generated with various peptide specific T cell yields (Fig. 1C). As shown, AT-SCT stimulation Downloaded from modifications, including N-terminal extensions of the agonist elicited a large expansion of T cells (2.4 3 108 T cells), with http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 2. Recognition of SCT and AT-SCT by soluble TCR multimers. (A) SCD were stained with mAbs to HLA-A2 (left panels) or pulsed with soluble 9-mer or 10-mer peptide and stained with HLA-A*0201/CMV pp65 NLV-specific TCR multimers (right panels). (B) NLV nonamers expressed by SCT and SCT dt molecules are equally recognized by TCR multimer (C), whereas NNLV decamer in AT-SCT or AT-SCT dt are not recognized by this TCR reagent. Dot plots are representative of cell populations expressing similar levels HLA-A2. The Journal of Immunology 5843 enrichment of the Ag-specific CD8+ T cell population (75.4% (23). Cell surface expression of the SCT and AT-SCT constructs tetramer+, 1.8 3 108 NLV-specific T cells). In contrast, the growth was determined using the HLA-A2–specific mAb (BB7.2) (Fig. 2, of the Ag-specific CD8+ T cell population was modest after SCT dot plots labeled HLA-A2). The HLA-A2/CMV pp65 TCR multi- stimulation (37.4% tetramer+, 1.6 3 107 NLV-specific T cells) mer stained 9-mer and 10-mer peptide-pulsed SCD-expressing (Fig. 1A, 1C). cells (Fig. 2A), as well as SCT-expressing cells (Fig. 2B). Sur- To further expand this observation, memory CD8+ T cells ob- prisingly, this CMV pp65 TCR multimer failed to stain the AT- tained from eight healthy CMV-seropositive donors (after NLV- SCT–expressing cells (Fig. 2C). Multiple factors, including pep- pulsed autologous mDC stimulation) were restimulated, in large- tide instability, could explain the lack of TCR multimer binding scale cultures, with SCT- or AT-SCT–expressing K562 cells; Ag- to AT-SCT. Next, we examined SCT dt constructs, because the specific yields and frequencies are shown in Fig. 1D and 1E. In all disulfide bond secures the peptide on SCT, making these mole- cases, AT-SCT–expressing cells were the superior agonist to cules unsusceptible to peptide exchange (22). As shown in Fig. stimulate NLV-specific CD8+ T cell growth (.10-fold increases in 2C, disulfide trap AT-SCT (AT-SCT dt) are not recognized by Ag-specific T cell yields over SCT, p = 0.0371, Fig. 1D). In ad- the TCR multimer, suggesting that NNLV instability is unlikely to dition, the frequency of Ag-specific CD8+ T cells (82.3% versus account for this lack of recognition. We previously demonstrated 63.8%, p = 0.0054, Fig. 1E) and fold increases in Ag-specific that peptide-competition assays allow one to examine occupancy T cells (563 6 160 versus 65 6 17, p = 0.0015, Fig. 1F) at day of the MHC peptide-binding groove (18, 22). As shown in Fig. 14 were significantly increased in the AT-SCT–stimulated culture 3A, exogenous competitor peptide could displace SCT (nonamer) compared with the SCT culture. and AT-SCT (decamer) peptides at similar concentrations, sug-

Additional AT-SCT constructs containing an amino acid residue gesting that linked peptides may display similar MHC occu- Downloaded from other than asparagine were tested in proliferation assays for their pancy. Importantly, and in accordance with proliferation results, ability to stimulate NLV-specific memory CD8+ T cells (see NLV-specific CD8+ T cells could recognize, equally well, in a Materials and Methods). Independent of the amino acid (Asn, Gly, standard [51Cr]-release assay, cells expressing SCT and AT-SCT or Tyr) identity, higher levels of proliferation were observed when (Fig. 3B) and SCT dt and AT-SCT dt (Fig. 3C), used as targets. memory NLV-specific T cells were stimulated with AT-SCT Altogether, these data strongly support the notion that p-MHC

compared with SCT (Supplemental Fig. 2). Altogether, these complexes in the AT-SCT format are structurally distinct from http://www.jimmunol.org/ results demonstrate that N-terminal extended NLV-peptide in the those on SCT. These changes may result in TCR-activation sig- SCT format (AT-SCT) promotes the numerical expansion (.500- nals that promote the numerical expansion of memory CD8+ fold) of memory CD8+ T cells upon restimulation. T cells. TCR multimers differ in their recognition of SCT and AT-SCT SCT and AT-SCT are ligands with distinct potencies To confirm the presentation of an amino terminal extended We investigated the early activation events of memory T cells (decamer) peptide by AT-SCT, HLA-A*0201 molecules were when stimulated by SCT- and AT-SCT–expressing cells. Upon isolated by immunoprecipitation from SCT- and AT-SCT–ex- TCR activation, memory CD8+ T cells upregulate CD69 and pressing K562 cells and subjected to amino acid sequencing by CD25 cell surface molecules, downregulate TCR expression, by guest on September 26, 2021 Edman degradation. The amino acid analysis confirms that the secrete IFN-g (and other cytokines), and undergo cell division. As SCT-expressing cells presented the NLV nonamer sequence, shown in Fig. 4A, the kinetics of CD69 and CD25 upregulation whereas the AT-SCT cells presented the NNLV decamer peptide were similar in SCT- and AT-SCT–stimulated T cells, suggesting (Supplemental Fig. 3). To further characterize p-MHC complexes, that early events associated with TCR triggering by SCT and AT- SCT and AT-SCT constructs expressed by K562 cells were probed SCT may be similar in these T cell populations. Importantly, upon using TCR multimers specific for HLA-A*0201/CMV pp65 NLV activation with either SCT or AT-SCT, these T cells displayed

FIGURE 3. Peptide competition and direct recogni- tion of SCT and AT-SCT in cytotoxicity assays. (A) An HLA-A2–restricted melanoma (G209-2M)-specific CTL line was used to monitor exogenous peptide binding to SCT and AT-SCT. Increasing concentrations of G209-2M competitor peptide were incubated with 51 Cr-labeled target cells expressing SCD, SCT, and AT- SCT. SCT and AT-SCT readily and similarly accepted exogenous peptide competitor at concentrations .0.1 nM, as shown by equivalent levels of lysis. Recognition of SCT and AT-SCT (B) and SCT dt and AT-SCT dt (C) by NLV-specific T cells was assessed using K562 cells expressing these constructs as targets in a standard [51Cr]-release assay. SCD-expressing cells pulsed with peptide were used as a positive control. The percentage of specific lysis (mean of triplicates) 6 1 SD is shown for each peptide concentration or E:T ratio. Spontane- ous lysis was ,10%. Results are representative of three experiments performed. 5844 SYNTHETIC AGONISTS FOR CD8+ T CELLS Downloaded from http://www.jimmunol.org/

FIGURE 4. Characterization of activation events upon stimulation with SCT and AT- SCT. (A) NLV-specific T cells (85% NLV/A*0201 tetramer+) were restimulated with SCT- or AT-SCT–expressing cells (1:1 ratio), and samples were collected at the indicated times. The expression of CD69 and CD25 on SCT- and AT-SCT–stimulated cells was determined by flow cytometry. Results are representative of two experiments. (B) Time course of ERK phos- phorylation (measured by flow cytometry) in NLV-specific (94% tetramer+) T cells after stimulation with K562-expressing SCT, AT-SCT, or media. by guest on September 26, 2021 Representative graph (at 10 min activation) is shown. Results are representative of three experiments performed. (C) T cells were activated as in (A), and TCR (CD3) downregulation was determined at the indicated times by flow cytometry (responding T cells were identified by IFN-g production). Results shown were obtained using a T cell population containing 52% NLV-specific T cells. Data are representative of results obtained with three donors. (D) Cytokine production in supernatants derived from NLV-specific T cells stimulated as indicated for 24 h. Cytokine quantitation was performed using a Milliplex cytokine kit. Results are mean of duplicates 6 1 SD and are representative of four experiments performed. (E) Summary of TCR downregulation observed at 24 h after stimulation with SCT and AT-SCT with multiple individuals in seven independent experiments. The p value was obtained using the Student t test; p values , 0.05 are significant. identical kinetics of ERK phosphorylation (Fig. 4B). In contrast, stimulation (98.8% NLV/A*0201 tetramer+) and stained with 24 h after activation, ∼80% of SCT-stimulated (Ag-specific) CFSE to monitor cell division. As shown in Fig. 5A, T cells from CD8+ T cells downregulate TCR (CD3) expression, whereas SCT- and AT-SCT–stimulated cultures enter mitosis at identical only 42% of AT-SCT–stimulated (Ag-specific) CD8+ T cells rates, as determined by CFSE content. In contrast, cell number and downregulate cell surface TCR (Fig. 4C). Interestingly, the per- viability were very different, as determined by manual cell counts centage of IFN-g–secreting cells was greater in cultures stimu- using trypan blue (Fig. 5B). Although cell numbers decreased in lated with SCT compared with AT-SCT at each time point SCT-stimulated T cell cultures, AT-SCT–stimulated T cells in- studied. As confirmation, we tested and found higher levels of creased in number, suggesting that the increased T cell growth IFN-g and other cytokines in supernatants obtained from SCT- observed in these cultures may result from decreased apoptosis. To stimulated T cells (Fig. 4C, 4D). The diminished TCR down- examine the fraction of T cells undergoing apoptosis during the regulation and IFN-g response after AT-SCT stimulation were proliferative response, NLV-specific T cells were stimulated with consistently observed in multiple experiments using memory SCT or AT-SCT. At various time points after stimulation, T cells NLV-specific CD8+ T cells derived from multiple individuals were stained with 7AAD and AnnexinFITC to monitor cell via- (Fig. 4E, data not shown). Altogether, these results suggest that bility. At 48 h postactivation, 79% of the SCT-stimulated T cells TCR signals transmitted by SCT and AT-SCT may be quantita- stained positive with annexin (16.4% annexin+/7AAD+), consis- tively different. tent with apoptotic cell death (Fig. 5C). In contrast, just 23% of the AT-SCT–stimulated T cells stained positive with annexin AT-SCT promotes proliferation with decreased cell death (6.1% annexin+/7AAD+). Therefore, despite similar proliferative We considered whether the striking numerical expansion of AT- rates, a smaller fraction of T cells undergo apoptosis in AT-SCT– SCT–stimulated T cells could be explained by either increased stimulated cultures compared with SCT-stimulated T cells, lead- cell proliferation or decreased cell death. To this end, NLV- ing to significantly higher yields of Ag-specific T cells 14 d after specific T cells were purified by cell sorting 14 d after mDC restimulation (Fig. 1). The Journal of Immunology 5845 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021 FIGURE 5. Decreased cell death rates in AT-SCT–stimulated T cells. (A) NLV-specific T cells (98.8% NLV/A*0201 tetramer+) were loaded with CFSE and restimulated with SCT- or AT-SCT–expressing cells; at the indicated times, they were harvested, counted, and analyzed by flow cytometry. The x-axis is the CFSE- staining intensity, whereas the y-axis represents the percentage of maximum cell number (% of Max). Equal numbers of events were collected on FSC/SSC-gated populations. Results are representative of two experiments performed. (B) Manual cell counts using trypan blue exclusion were performed to determine viability in culturesshownin(A). Bars represent cell counts obtained at the indicated time point from SCT- or AT-SCT–stimulated cultures. Results are representative of two experiments performed. (C) NLV-specific T cells (96% NLV/A*0201 tetramer+) were restimulated with SCT- and AT-SCT–expressing cells, and apoptotic T cells were identified on the indicated days by staining with 7AAD and AnnexinFITC. Camptothecin-treated (10 mM) T cells were used as positive controls to set up dot plot quadrants (data not shown). Percentages represent annexin+/annexin+-7AAD+ T cell populations. Media indicates T cell cultures in Stemline-5 in the absence of stimulation. Equal numbers of events were collected on the FSC/SSC-gated populations. Results are representative of three experiments performed.

AT-SCT–stimulated T cells display “early” effector CD8+ A*0201+ target cells (Fig. 6D). Finally, to determine whether AT- characteristics SCT stimulation resulted in the preferential expansion of selected Large-scale cultures of AT-SCT–stimulated T cells, obtained from clones, the TCR repertoire was characterized using a Vb five of the eight donors shown in Fig. 1E and 1F, were studied to clonality-detection assay. As shown in Fig. 6E and 6F, AT-SCT– assess functional and phenotypic traits (25, 29). Cytokine (IFN-g, expanded NLV-specific T cells used a diverse set of TCR Vb IL-2, TNF-a, MIP-1a) production by AT-SCT–stimulated NLV- regions, including Vb subsets (BV8, BV13, BV14, BV20) fre- specific T cells was evaluated by intracellular staining using quently identified in healthy CMV-seropositive individuals (35). Collectively, these results show that stimulation of memory human polychromatic flow cytometry. In T cell cultures obtained from + . + CD8 T cells with AT-SCT results in a population of polyfunc- five healthy individuals, 80% of NLV-specific CD8 T cells were + polyfunctional (produced more than one cytokine), and .25% of tional and high avidity Ag-specific CD8 T cells. these cells were positive for all four functions (Fig. 6A). Pheno- typically, .80% of the expanded cells expressed CD28, and 22– Melanoma-specific T cells are efficiently expanded with 46% were positive for CD27 (Fig. 6B). All cells were positive for AT-SCT ligands CD45RO and effector molecules granzyme A, granzyme B, and To investigate whether AT-SCT ligands in other antigenic systems perforin (Fig. 6B). Thus, despite the large numerical expansion share the structural and functional features described in the undergone by AT-SCT–stimulated cells, these retain an “early” CMV NLV-peptide system, the melanoma gp100-derived antigenic effector memory phenotype (Supplemental Fig. 4). Expanded peptide G209-2M was studied (36). An HLA-A*0201/G209-2M– T cells displayed high avidity for Ag (EC50 , 100 pM) when specific TCR multimer reagent, similar to the one used in the tested in a standard 4-h [51Cr] assay using peptide-pulsed T2 NLV-peptide system, was used to probe the p-MHC complexes target cells (Fig. 6C), and they recognized CMV-infected HLA- expressed by SCT (IMD nonamer), as well as two AT-SCT 5846 SYNTHETIC AGONISTS FOR CD8+ T CELLS Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021 FIGURE 6. Functional and phenotypic characteristics of AT-SCT–stimulated NLV-specific T cells. (A) NLV-specific T cells obtained from five individual donors (D1–D5) after AT-SCT restimulation were functionally and phenotypically characterized. Ag-specific frequencies, as determined by NLV/A*0201 tetramer staining, range from 75 to 96% of CD8+ T cells. To evaluate polyfunctionality, T cells were restimulated with SCT for 16 h in the presence of brefeldin A, permeabilized, and stained for cytokine expression. Total frequency of Ag-specific T cells represents the summation of T cell frequencies within each unique cytokine combination (data not shown); the responses are grouped according to the number of cytokines produced. Pie charts depict polyfunctional profiles, with responses grouped according to the number of functions (IFN-g, IL-2, TNF-a, and MIP-1b). (B) Cells characterized in (A) were stained with NLV/A*0201 tetramer and Abs directed at the indicated cell surface markers. A total of 25,000 events was collected on NLV/A*0201 tetramer+ gated cells; the percentages of T cells for the indicated cell surface molecules within the tetramer+ population are reported. (C) Avidity of NLV- specific T cells was determined in a standard 4-h [51Cr]-release assay using peptide titrations and T2 cells as targets. The percentage of specific lysis (mean of triplicates) is shown for each peptide concentration. SD was ,20% and is omitted for clarity; spontaneous lysis was ,5%. Results are representative of two experiments performed. (D) To demonstrate recognition of native CMV Ag by AT-SCT–expanded T cells, T98G glioma cells were infected with CMV (27) for 72 h or pulsed with peptide, labeled with [51Cr], and used as targets in a standard assay. The percentage of specific lysis (mean of triplicates) 6 1 SD is shown for each target. Spontaneous lysis was ,10%. Results are representative of three experiments performed. (E and F) After DC stimulation, NLV-specific T cells were purified by cell sorting using NLV/A*0201 tetramers. The sorted populations were stimulated with AT-SCT in the presence of IL- 2, as described in Materials and Methods, and the frequency of NLV/A*0201 tetramer CD8+ T cells on day 14 after restimulation is shown as dot plots. RNA was isolated from T cells, and TCR Vb usage was determined as described in Materials and Methods. A similar analysis was performed with CD8+ T cells isolated from fresh PBMC prior to Ag stimulation. Top panel, TCR Vb usage in CD8+ T cell population prior to Ag stimulation. Bottom panel, TCR Vb usage in NLV-specific T cells after AT-SCT stimulation. The results obtained with two representative donors are shown.

(NIMD and TIMD decamer)-expressing K562 cell lines. Similarly the frequency of peptide-specific CD8+ T cells by flow cytometry to our observations in the CMV Ag system, the G209-2M TCR (Fig. 7B). Memory (G209-2M Ag-specific) T cells restimulated multimer reagent failed to stain either AT-SCT, whereas it dem- with AT-SCT showed significant growth over 10 d, whereas the onstrated positive staining on SCT (Fig. 7A). The functional SCT-stimulated memory T cells displayed minimal growth and consequences of AT-SCT stimulation were examined using CD8+ dramatically decreased in frequency (Fig. 7B, 7C). As shown for a T cells obtained from HLA-A*0201+ melanoma patients after representative patient, cultures started with 5 3 106 cells (0.38 3 autologous mDC vaccination. CD8+ T cells purified from pe- 106 G209+) yielded 1.4 3 108 (6.1 3 107 G209+) and 4.9 3 108 ripheral blood were stimulated with G209-2M peptide-pulsed (1.8 3 108 G209+) total T cells after stimulation with AT-SCT autologous mDC for 14 d and restimulated with SCT- or AT- NIMD and TIMD, respectively (Fig. 7C). In contrast, cells SCT–expressing K562 cells. Live cell growth kinetics were de- stimulated with SCT decreased in frequency, suggesting clonal termined by manual counting using trypan blue exclusion at ini- deletion. AT-SCT yields represent Ag-specific T cell increases of tiation and termination of the culture; the surviving cells were 10- and 30-fold over cultures stimulated with SCT (Fig. 7D). stained with anti-CD8 and G209-2M/A*0201 tetramer to assess Similar to the CMV model system, the melanoma G209-2M– The Journal of Immunology 5847

FIGURE 7. AT-SCT promotes the expansion of melanoma-specific T cells. (A) K562 cells expressing SCT and AT-SCT were stained with mAbs to HLA-A2 (BB7.2, left panels) or TCR multimers specific for HLA-A*0201/G209-2M complexes (right panels). IMD nonamers expressed by SCT are recognized by the TCR reagent, whereas the NIMD and TIMD deca- mers on AT-SCT lack reactivity with TCR re- agent, despite similar levels of HLA-A2 class I. Results are representative of two experiments performed. (B) Frequencies of G209-2M–spe- cific T cells after DC stimulation and SCT (IMD) and AT-SCT (NIMD and TIMD) re- stimulation, as determined by flow cytometry

using G209-2M/A*0201 tetramers. Results are Downloaded from representative of those obtained with three patients. (C and D) DC-activated T cells (5 3 106 total T cells, 0.38 3 106 G209-2M specific) were restimulated with SCT- or AT-SCT– expressing K562 cells at a 1:1 ratio, and cul- tures were maintained as described in Materi- als and Methods. Ag-specific yields and folds http://www.jimmunol.org/ are shown for a representative donor. Yields and folds [indicated above the bars in (D)] were calculated as described in Fig. 1. (E and F) G209-2M–specific T cells expanded with AT- SCT were tested for their ability to recognize native gp100 Ag on human melanoma cells. The results are shown for melanoma cell lines DM6 (HLA-A2+/gp100+)(E) and A375 (HLA- A2+/gp1002)(F); cell lines were tested as tar- gets in the presence (s) or absence (d)of by guest on September 26, 2021 G209-2M peptide. Percentage of specific lysis (mean of triplicates) 6 1 SD is shown; spon- taneous lysis was ,10%. Results are repre- sentative of six experiments performed.

specific T cells exhibited slower TCR downregulation after AT- expansion of memory CD8+ T cells elicited by natural infection SCT activation compared with SCT activation, and the expanded (CMV), as well as therapeutic vaccination (melanoma gp100 Ag). T cells killed human HLA-A*0201+/gp100+ melanoma cells The SCT scaffold for peptide presentation was used for two in vitro (Fig. 7E). reasons. First, SCT ensures structural integrity and assumes that each p-MHC is equivalent. Second, SCT constructs can be ex- Discussion clusively presented by mutant cell lines (such as K562), which lack Using HLA-A*0201–restricted viral and self-Ags (CMV pp65 endogenous MHC molecules, minimizing peptide exchange and NLV peptide and melanoma gp100 G209-2M peptide, respec- allowing a uniform expression of a single p-MHC species. Previous tively), we provide evidence that N-terminal extended peptides studies demonstrated that single-chain p-MHC can induce pro- (decamers), when presented as single-chain MHC class I trimers liferation, inhibit CTL-mediated cytotoxicity, or induce cell death (AT-SCT), are highly stimulatory for memory human CD8+ (21, 37, 38). Structural features of single-chain p-MHC, such as T cells, resulting in the growth of a clonally diverse population of short versus long linkers or soluble versus cell bound, as well as Ag-specific T cells with polyfunctional traits. The expanded T cell the differentiation status of T cells, play a role in determining the population displays phenotypic traits of early effector CD8+ outcome of single-chain p-MHC/T cell interaction (21, 37, 38). In T cells, CD45RA2/CD28+/CD27high/low/CCR5+, and retains high this study, we find that insertion of an amino acid residue at the N avidity for Ag (29). Our results demonstrate that activation with terminus of antigenic peptide of a single-chain trimer (AT-SCT) AT-SCT leads to increases in the frequency and numbers of Ag- creates a novel structure that promotes ex vivo numerical expan- specific CD8+ T cells. In both viral and self-Ag systems, we sion of memory CD8+ T cell with maintenance of functional ca- routinely observe .400-fold increases in Ag-specific T cell pacity (Fig. 6). We report these results in two clinically relevant numbers after AT-SCT restimulation. Thus, AT-SCT defines antigenic systems: CMV pp65 and melanoma gp100 (39, 40). a novel p-MHC entity that can act as a potent agonist for the However, in our experience, random insertion of any amino acid 5848 SYNTHETIC AGONISTS FOR CD8+ T CELLS residue at the N-terminal position may not always be sufficient to may induce TCR signals that are qualitatively different from those create a stimulatory AT-SCT. For example, addition of K (natu- delivered by canonical nonamer SCT (Figs. 4, 5). Altogether, it is rally occurring amino acid residue) or N to the N terminus of the tempting to speculate that AT-SCT represent unique p-MHC en- influenza A matrix (MP 58-66, HLA-A*0201–restricted) nonamer tities that allow the segregation of TCR proliferative and death peptide failed to confer improved agonist activity to the flu M1 signals as a result of a change in the strength of the signal. De- AT-SCT relative to the SCT (B.M. Carreno and M. Becker-Hapak, veloping this thesis postulates that TCR signals for cytokine unpublished observations). Structural data on the AT-SCT entities production can be clearly segregated from signals that promote described in this article should be instructive in modeling pre- cell death. Indeed, a previous study demonstrated that APL can dictions for additional peptide Ags in the context of HLA-A2, as selectively induce expression of Fas ligand (CD95L) and mediate well as other HLA class I alleles. apoptosis via Fas signaling without concomitant cytokine pro- AT-SCT stimulation of memory CD8+ T cells leads to slower duction (44). cell surface TCR (CD3) downregulation, decreased cytokine In conclusion, AT-SCT expressed on MHC class I/II-null K562 synthesis, and reduced RICD. These features are discernible cells can stimulate .400-fold expansion of memory human CD8+ quantitative differences from canonical nonamer SCT stimulation T cells over a 10–14-d culture period. The expanded T cell pop- and may define a novel mechanism of T cell activation. In con- ulation is clonally diverse, polyfunctional, and highly avid in the trast, other early signaling events, such as ERK phosphorylation recognition of native Ag. Our study demonstrates the potential use and upregulation of activation markers, such as CD25 and CD69, of AT-SCT as a novel synthetic agonist for the in vitro numerical are identical in AT-SCT– and SCT-stimulated T cells. These expansion of T cells for adoptive cell therapy and suggests new

findings suggest that AT-SCT represent p-MHC structures distinct opportunities to engineer potent agonistic p-MHC constructs for Downloaded from from most APL, which induce a state of partial activation (9); applications in cancer and infectious diseases. indeed, AT-SCT induces full activation based on cell cycle entry kinetics with avoidance of apoptotic cell death. This is particularly Acknowledgments striking in the self-peptide G209-2M system (Fig. 7B), in which We thank Dr. Dong Yu for providing AD169 CMV, Drs. Janet Connolly and restimulation with AT-SCT leads to increases in the frequency and Dave Freemont for helpful discussions, and Drs. Paul Allen and Joaquin number of Ag-specific T cells, whereas SCT leads to clonal de- Madrenas for critical reading of the manuscript. Flow cytometry was per- http://www.jimmunol.org/ + letion, because only ∼1% tetramer T cells remain after 14 d of formed at the Siteman Cancer Center High-Speed Cell Sorter Core Facility. culture. In the CMV pp65 system, activation with SCT promotes limited growth, generally without enrichment of tetramer fre- Disclosures quencies, whereas AT-SCT promotes marked growth and enrich- The authors have no financial conflicts of interest. ment. Thus, native Ag leads to RICD of memory T cells, whereas AT-SCT promotes T cell expansion to increase the memory pool. These findings raise the interesting possibility that native Ag may References not constitute the optimal immunogen for booster for- 1. Rudolph, M. G., R. L. Stanfield, and I. A. Wilson. 2006. How TCRs bind MHCs, peptides, and coreceptors. Annu. Rev. Immunol. 24: 419–466. + by guest on September 26, 2021 mulations designed to promote a therapeutic CD8 response or 2. Snow, A. L., P. Pandiyan, L. Zheng, S. M. Krummey, and M. J. Lenardo. 2010. long-lived memory response (41). As a corollary, native agonist The power and the promise of restimulation-induced cell death in human im- peptide would be required for the effective priming of naive mune diseases. Immunol. Rev. 236: 68–82. + 3. Wells, A. D. 2009. New insights into the molecular basis of T cell anergy: anergy CD8 T cells specific for low-affinity self-peptides, such as the factors, avoidance sensors, and epigenetic imprinting. J. Immunol. 182: 7331–7341. melanocyte-derived proteins; in contrast, AT-SCT would be inef- 4. Zhang, N., and M. J. Bevan. 2011. CD8(+) T cells: foot soldiers of the immune + system. Immunity 35: 161–168. fective for CD8 T cell priming against self or viral Ags. 5. Schodin, B. A., T. J. Tsomides, and D. M. Kranz. 1996. Correlation between the Structure–function studies (42) examining A6 TCR binding number of T cell receptors required for T cell activation and TCR-ligand affinity. to HLA-A2/Tax peptide variants conclude that almost identical p- Immunity 5: 137–146. 6. Davis, M. M., J. J. Boniface, Z. Reich, D. Lyons, J. Hampl, B. Arden, and MHC complexes can induce different T cell signals, as assessed Y. Chien. 1998. Ligand recognition by alpha beta T cell receptors. Annu. Rev. by cytokine release, TCR downregulation, and CTL-mediated Immunol. 16: 523–544. cytotoxicity. The differential recognition of SCT and AT-SCT p- 7. Huang, J., V. I. Zarnitsyna, B. Liu, L. J. Edwards, N. Jiang, B. D. Evavold, and C. Zhu. 2010. The kinetics of two-dimensional TCR and pMHC interactions MHC complexes by TCR multimers suggests that accommoda- determine T-cell responsiveness. Nature 464: 932–936. tion of nonamer- and decamer-linked peptides in the MHC-binding 8. Bridgeman, J. S., A. K. Sewell, J. J. Miles, D. A. Price, and D. K. Cole. 2012. Structural and biophysical determinants of ab T-cell recognition. Im- groove may be distinct. Multiple factors could account for the munology 135: 9–18. lack of recognition of AT-SCT by TCR multimers, including a 9. Sloan-Lancaster, J., and P. M. Allen. 1996. Altered peptide ligand-induced conformational change in the peptide, steric hindrance due to the partial T cell activation: molecular mechanisms and role in T cell biology. Annu. Rev. Immunol. 14: 1–27. additional N-terminal residue, and/or peptide instability. Although 10. Madrenas, J., and R. N. Germain. 1996. Variant TCR ligands: new insights into we cannot formally rule out any of these possibilities, we think the molecular basis of antigen-dependent signal transduction and T-cell activa- that peptide instability is unlikely, because AT-SCT in the secure tion. Semin. Immunol. 8: 83–101. 11. Tangri, S., G. Y. Ishioka, X. Huang, J. Sidney, S. Southwood, J. Fikes, and disulfide trap format are not recognized by the TCR multimer A. Sette. 2001. Structural features of peptide analogs of human histocompati- (Fig. 2C) but are recognized by CTL (Fig. 3C). Because of the bility leukocyte antigen class I epitopes that are more potent and immunogenic than wild-type peptide. J. Exp. Med. 194: 833–846. close nature of the P1 pocket in HLA-A2, the decamer peptide 12. Abdul-Alim, C. S., Y. Li, and C. Yee. 2010. Conditional superagonist CTL may adopt a bulge conformation with a change in peptide anchor ligands for the promotion of tumor-specific CTL responses. J. Immunol. 184: residue at P2. This proposed scenario is similar to that described 6514–6521. 13. Inaba, K., and R. M. Steinman. 1984. Resting and sensitized T lymphocytes for the HLA-A2–restricted 10-mer MART-1 peptides, in which exhibit distinct stimulatory (antigen-presenting cell) requirements for growth central bulging of a long peptide allows for its accommodation and lymphokine release. J. Exp. Med. 160: 1717–1735. within the peptide-binding groove (43). Regardless of the mech- 14. Croft, M., L. M. Bradley, and S. L. Swain. 1994. Naive versus memory CD4 T cell response to antigen. Memory cells are less dependent on accessory cell anism, AT-SCT stimulation results in productive TCR engage- costimulation and can respond to many antigen-presenting cell types including ment, as demonstrated by numerical T cell expansion (Figs. 1, 7) resting B cells. J. Immunol. 152: 2675–2685. 15. Russell, J. H., C. L. White, D. Y. Loh, and P. Meleedy-Rey. 1991. Receptor- and recognition of AT-SCT–expressing targets in CTL-mediated stimulated death pathway is opened by antigen in mature T cells. Proc. Natl. cytotoxicity assays (Fig. 3B, data not shown). However, AT-SCT Acad. Sci. USA 88: 2151–2155. The Journal of Immunology 5849

16. Lenardo, M. J. 1991. Interleukin-2 programs mouse alpha beta T lymphocytes 31. Thomas, A. K., M. V. Maus, W. S. Shalaby, C. H. June, and J. L. Riley. 2002. A for apoptosis. Nature 353: 858–861. cell-based artificial antigen-presenting cell coated with anti-CD3 and CD28 17. Iezzi, G., K. Karjalainen, and A. Lanzavecchia. 1998. The duration of antigenic antibodies enables rapid expansion and long-term growth of CD4 stimulation determines the fate of naive and effector T cells. Immunity 8: 89–95. T lymphocytes. Clin. Immunol. 105: 259–272. 18. Yu, Y. Y., N. Netuschil, L. Lybarger, J. M. Connolly, and T. H. Hansen. 2002. 32. Suhoski, M. M., T. N. Golovina, N. A. Aqui, V. C. Tai, A. Varela-Rohena, Cutting edge: single-chain trimers of MHC class I molecules form stable M. C. Milone, R. G. Carroll, J. L. Riley, and C. H. June. 2007. Engineering structures that potently stimulate antigen-specific T cells and B cells. J. Immunol. artificial antigen-presenting cells to express a diverse array of co-stimulatory 168: 3145–3149. molecules. Mol. Ther. 15: 981–988. 19. Hansen, T. H., J. M. Connolly, K. G. Gould, and D. H. Fremont. 2010. Basic and 33. Zhang, H., K. M. Snyder, M. M. Suhoski, M. V. Maus, V. Kapoor, C. H. June, translational applications of engineered MHC class I proteins. Trends Immunol. and C. L. Mackall. 2007. 4-1BB is superior to CD28 costimulation for gener- 31: 363–369. ating CD8+ cytotoxic lymphocytes for adoptive immunotherapy. J. Immunol. 20. Mitaksov, V., S. M. Truscott, L. Lybarger, J. M. Connolly, T. H. Hansen, and 179: 4910–4918. D. H. Fremont. 2007. Structural engineering of pMHC reagents for T cell vac- 34. Hemmer, B., I. Stefanova, M. Vergelli, R. N. Germain, and R. Martin. 1998. cines and diagnostics. Chem. Biol. 14: 909–922. Relationships among TCR ligand potency, thresholds for effector function 21. Choudhuri, K., D. Wiseman, M. H. Brown, K. Gould, and P. A. van der Merwe. elicitation, and the quality of early signaling events in human T cells. J. 2005. T-cell receptor triggering is critically dependent on the dimensions of its Immunol. 160: 5807–5814. peptide-MHC ligand. Nature 436: 578–582. 35. Trautmann, L., M. Rimbert, K. Echasserieau, X. Saulquin, B. Neveu, 22. Truscott, S. M., X. Wang, L. Lybarger, W. E. Biddison, C. McBerry, J. Dechanet, V. Cerundolo, and M. Bonneville. 2005. Selection of T cell clones J. M. Martinko, J. M. Connolly, G. P. Linette, D. H. Fremont, T. H. Hansen, and expressing high-affinity public TCRs within Human cytomegalovirus-specific B. M. Carreno. 2008. Human major histocompatibility complex (MHC) class I CD8 T cell responses. J. Immunol. 175: 6123–6132. molecules with disulfide traps secure disease-related antigenic peptides and 36. Parkhurst, M. R., M. L. Salgaller, S. Southwood, P. F. Robbins, A. Sette, exclude competitor peptides. J. Biol. Chem. 283: 7480–7490. S. A. Rosenberg, and Y. Kawakami. 1996. Improved induction of melanoma- 23. Zhu, X., H. J. Belmont, S. Price-Schiavi, B. Liu, H. I. Lee, M. Fernandez, reactive CTL with peptides from the melanoma antigen gp100 modified at HLA- R. L. Wong, J. Builes, P. R. Rhode, and H. C. Wong. 2006. Visualization of p53 A*0201-binding residues. J. Immunol. 157: 2539–2548. (264-272)/HLA-A*0201 complexes naturally presented on tumor cell surface by 37. Cebecauer, M., P. Guillaume, P. Hoza´k, S. Mark, H. Everett, P. Schneider, and a multimeric soluble single-chain T cell receptor. J. Immunol. 176: 3223–3232. I. F. Luescher. 2005. Soluble MHC-peptide complexes induce rapid death of Downloaded from 24. Carreno, B. M., M. Becker-Hapak, and G. P. Linette. 2009. CD40 regulates CD8+ CTL. J. Immunol. 174: 6809–6819. human dendritic cell-derived IL-7 production that, in turn, contributes to CD8(+) 38. Samanta, D., G. Mukherjee, U. A. Ramagopal, R. J. Chaparro, S. G. Nathenson, T-cell antigen-specific expansion. Immunol. Cell Biol. 87: 167–177. T. P. DiLorenzo, and S. C. Almo. 2011. Structural and functional characteriza- 25. Precopio, M. L., M. R. Betts, J. Parrino, D. A. Price, E. Gostick, tion of a single-chain peptide-MHC molecule that modulates both naive and D. R. Ambrozak, T. E. Asher, D. C. Douek, A. Harari, G. Pantaleo, et al. 2007. activated CD8+ T cells. Proc. Natl. Acad. Sci. USA 108: 13682–13687. Immunization with vaccinia virus induces polyfunctional and phenotypically 39. Peggs, K., S. Verfuerth, A. Pizzey, J. Ainsworth, P. Moss, and S. Mackinnon. distinctive CD8(+) T cell responses. J. Exp. Med. 204: 1405–1416. 2002. Characterization of human cytomegalovirus peptide-specific CD8(+) T-

26. Linette, G. P., D. Zhang, F. S. Hodi, E. P. Jonasch, S. Longerich, C. P. Stowell, cell repertoire diversity following in vitro restimulation by antigen-pulsed den- http://www.jimmunol.org/ I. J. Webb, H. Daley, R. J. Soiffer, A. M. Cheung, et al. 2005. Immunization dritic cells. Blood 99: 213–223. using autologous dendritic cells pulsed with the melanoma-associated antigen 40. Yang, S., G. P. Linette, S. Longerich, and F. G. Haluska. 2002. Antimelanoma gp100-derived G280-9V peptide elicits CD8+ immunity. Clin. Cancer Res. 11: activity of CTL generated from peripheral blood mononuclear cells after stim- 7692–7699. ulation with autologous dendritic cells pulsed with melanoma gp100 peptide 27. Yu, D., G. A. Smith, L. W. Enquist, and T. Shenk. 2002. Construction of a self- G209-2M is correlated to TCR avidity. J. Immunol. 169: 531–539. excisable bacterial artificial chromosome containing the human cytomegalovirus 41. Kenter, G. G., M. J. Welters, A. R. Valentijn, M. J. Lowik, D. M. Berends-van genome and mutagenesis of the diploid TRL/IRL13 gene. J. Virol. 76: 2316– der Meer, A. P. Vloon, F. Essahsah, L. M. Fathers, R. Offringa, J. W. Drijfhout, 2328. et al. 2009. Vaccination against HPV-16 oncoproteins for vulvar intraepithelial 28. Gras, S., X. Saulquin, J. B. Reiser, E. Debeaupuis, K. Echasserieau, neoplasia. N. Engl. J. Med. 361: 1838–1847. A. Kissenpfennig, F. Legoux, A. Chouquet, M. Le Gorrec, P. Machillot, et al. 42. Ding, Y. H., B. M. Baker, D. N. Garboczi, W. E. Biddison, and D. C. Wiley. 2009. Structural bases for the affinity-driven selection of a public TCR against 1999. Four A6-TCR/peptide/HLA-A2 structures that generate very different

a dominant human cytomegalovirus epitope. J. Immunol. 183: 430–437. T cell signals are nearly identical. Immunity 11: 45–56. by guest on September 26, 2021 29. Takata, H., and M. Takiguchi. 2006. Three memory subsets of human CD8+ 43. Borbulevych, O. Y., F. K. Insaidoo, T. K. Baxter, D. J. Powell, Jr., L. A. Johnson, T cells differently expressing three cytolytic effector molecules. J. Immunol. N. P. Restifo, and B. M. Baker. 2007. Structures of MART-126/27-35 Peptide/ 177: 4330–4340. HLA-A2 complexes reveal a remarkable disconnect between antigen structural 30.Maus,M.V.,A.K.Thomas,D.G.Leonard,D.Allman,K.Addya homology and T cell recognition. J. Mol. Biol. 372: 1123–1136. K. Schlienger, J. L. Riley, and C. H. June. 2002. Ex vivo expansion of polyclonal 44. Combadie`re, B., C. Reis e Sousa, R. N. Germain, and M. J. Lenardo. 1998. and antigen-specific cytotoxic T lymphocytes by artificial APCs expressing Selective induction of apoptosis in mature T lymphocytes by variant T cell re- ligands for the T-cell receptor, CD28 and 4-1BB. Nat. Biotechnol. 20: 143–148. ceptor ligands. J. Exp. Med. 187: 349–355.