Download Functional Avidity–Driven Activation

Functional Avidity−Driven
Activation-Induced Cell Death Shapes CTL
Immunodominance
This information is current as
of June 15, 2017.
Silvia Dalla Santa, Anna Merlo, Sara Bobisse, Elisa
Ronconi, Daniela Boldrin, Gabriella Milan, Vito Barbieri,
Oriano Marin, Antonella Facchinetti, Giovanni Biasi,
Riccardo Dolcetti, Paola Zanovello and Antonio Rosato
Supplementary
Material
References
Subscription
Permissions
Email Alerts
http://www.jimmunol.org/content/suppl/2014/09/22/jimmunol.130320
3.DCSupplemental
This article cites 42 articles, 23 of which you can access for free at:
http://www.jimmunol.org/content/193/9/4704.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
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 © 2014 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
J Immunol 2014; 193:4704-4711; Prepublished online 22
September 2014;
doi: 10.4049/jimmunol.1303203
http://www.jimmunol.org/content/193/9/4704
The Journal of Immunology
Functional Avidity–Driven Activation-Induced Cell Death
Shapes CTL Immunodominance
Silvia Dalla Santa,* Anna Merlo,* Sara Bobisse,† Elisa Ronconi,‡ Daniela Boldrin,*
Gabriella Milan,x Vito Barbieri,{ Oriano Marin,‖ Antonella Facchinetti,{
Giovanni Biasi,# Riccardo Dolcetti,** Paola Zanovello,*,{ and Antonio Rosato*,{
T
he immune response usually focuses on one or just a few
antigenic epitopes in a hierarchic and highly reproducible
manner, a phenomenon termed “immunodominance” (ID)
(1, 2). This can be influenced at every step of the immune response, ranging from epitope availability, as in the case of viral
infections that produce antigenic epitopes with a specific timetable
(3, 4), to Ag processing and presentation (5, 6), and, finally, to
T–T (7) and T–APC interactions (8).
An essential condition to enter the hierarchy is an efficient
processing of the antigenic determinants by the immunoproteasome. This step depends on the epitope binding affinity for the
transporter associated with Ag processing and has to achieve
a well-defined processing threshold (1, 6). The subsequent assembly of peptide–MHC complexes is critically influenced by the
epitope binding affinity to MHC, with the immunodominant
determinants (IDDs) usually having a higher affinity for the
complex (9) than the subdominant determinants (SDDs) (10).
*Veneto Institute of Oncology, 35128 Padua, Italy; †Ludwig Center for Cancer Research, University of Lausanne, Biopôle III, 1066 Epalinges, Lausanne, Switzerland;
‡
Excellence Centre for Research, Transfer, and High Education, University of Florence, 50139 Florence, Italy; xDepartment of Medicine, University of Padua, 35128
Padua, Italy; {Department of Surgery, Oncology, and Gastroenterology, University of
Padua, 35128 Padua, Italy; ‖Department of Biomedical Sciences, University of
Padua, 35131 Padua, Italy; #Department of Molecular Pathology, University of Marche, 60126 Ancona, Italy; and **National Cancer Institute, 33081 Aviano, Italy
Received for publication November 27, 2013. Accepted for publication August 21,
2014.
This work was supported by the Italian Association for Cancer Research (IG-13121
and Special Program Molecular Clinical Oncology 5 per Mille ID 10016 [to A.R.]
and IG-14287 [to R.D.]) and the University of Padua “Progetti Strategici di Ateneo
2011” (to A.R.).
Address correspondence and reprint requests to Dr. Antonio Rosato, Department of
Surgery, Oncology, and Gastroenterology, University of Padua, Via Gattamelata 64,
I-35128 Padua, Italy. E-mail address: [email protected].
The online version of this article contains supplemental material.
Abbreviations used in this article: ACT, adoptive cell therapy; AICD, activationinduced cell death; B6, C57BL/6; ID, immunodominance; IDD, immunodominant
determinant; MLPC, mixed leukocyte peptide culture; Mo-MSV/Mo-MuLV,
Moloney-murine sarcoma/leukemia virus; SDD, subdominant determinant; TIL, infiltrating T lymphocyte.
Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1303203
The frequency of responsive T cells with a TCR specific for
a given peptide–MHC complex is another key point for ID. In fact,
during a cellular immune response to a pathogen, the CD8+ T cell
population undergoes a strong selection for a highly restricted TCR
repertoire (11, 12), to limit potential autoimmune reactions (13).
Then, the extent of recruitment, the length of the precursor expansion phase (14), and the proliferative potential of each specific
CTL clone (13) are involved in the generation of the response
hierarchy. In turn, all of these factors potentially can be influenced
by the functional avidity of Ag-specific T cells, a parameter that
describes how well a T cell responds to Ag stimulation (15, 16).
Finally, the interactions of Ag-specific T cells with APCs or other
T cell subpopulations also can contribute to the establishment of ID
hierarchies. In this regard, the immunodomination phenomenon
describes the T–T cell competition for access to the same APC (8),
which leads to the inhibition or the suppression of the T cell response to a given Ag by other T cells and may occur when they
share the same APC and the APCs are scarce (8).
In this study, we describe a novel mechanism of ID that characterizes the immune response to a retroviral complex (Moloneymurine sarcoma/leukemia virus [Mo-MSV/Mo-MuLV]), which
gives rise to sarcomas rapidly undergoing spontaneous regression
mediated by virus-specific CTLs (17). Previously, we identified the
apparent IDD and SDD antigenic determinants in Gag85–93 and
Env189–196 epitopes, respectively, in the H-2b mouse strain (18).
We now report that, both in vitro and in vivo, SDD-specific CTLs
exhibit a much greater Ag avidity than do the IDD-recognizing
counterparts, and they undergo apoptosis due to Ag overload and
supraoptimal TCR engagement. These features prevent their expansion, ultimately allowing ID of the less-avid Gag-specific
population. Therefore, we propose avidity-dependent hyperactivation-induced cell death as a novel mechanism in the establishment of the ID hierarchy of CD8+ T cells.
Materials and Methods
Mice
Six- to eight-week-old female C57BL/6 (B6) mice (H-2b) (Charles River
Laboratories) were housed in a specific pathogen free animal facility.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Immunodominance is a complex phenomenon that relies on a mere numerical concept, while being potentially influenced at every
step of the immune response. We investigated the mechanisms leading to the establishment of CTL immunodominance in
a retroviral model and found that the previously defined subdominant Env-specific CD8+ T cells are endowed with an unexpectedly higher functional avidity than is the immunodominant Gag-recognizing counterpart. This high avidity, along with the Env Ag
overload, results in a supraoptimal TCR engagement. The overstimulation makes Env-specific T lymphocytes more susceptible to
apoptosis, thus hampering their expansion and leading to an unintentional “immune kamikazing.” Therefore, Ag-dependent,
hyperactivation-induced cell death can be regarded as a novel mechanism in the establishment of the immunodominance that
restrains and opposes the expansion of high-avidity T cells in favor of lower-affinity populations. The Journal of Immunology,
2014, 193: 4704–4711.
The Journal of Immunology
Procedures involving animals and their care conformed with institutional
guidelines (D.L. 116/92 and subsequent implementing notes), and experimental protocols (1186/05; 1130/08) were approved by the Italian Ministry of Health.
Virus preparation and administration
The Mo-MSV/Mo-MuLV cell extract was prepared as previously described
(18). Adult B6 mice were injected i.m. in the hind region with 150 ml cellfree preparation.
Synthetic peptides and preparation/synthesis of MHC–peptide
tetrameric complexes
Gag85–93 [H-2Db–restricted, CCLCLTVFL (18)] and Env189–196 [H-2Kb–
restricted, SSWDFITV (18)] peptides and the relative controls (NP366–374
from the influenza A virus, H-2Db restricted, ASNENMETM and b-gal96–103,
H-2Kb restricted, DAPIYTNV) were obtained from Tecnogen. A Gagmodified peptide (19) was used only for tetramer preparation and was
synthesized at the Centro Ricerca Interdipartimentale Biotecnologie Innovative of Padua University. Soluble tetramers were produced as previously described (20).
Tumor cell lines
CTL clones, mixed leukocyte peptide cultures, and infiltrating
T lymphocyte isolation
CTL clones and mixed leukocyte peptide cultures (MLPCs) were obtained
as previously described (18). Infiltrating T lymphocytes (TILs) were isolated from tumor masses as reported (21), without Ficoll-Paque (GE
Healthcare) treatment.
and FITC-conjugated annexin (Annexin-V-FLUOS Staining Kit; Roche).
Data analysis was carried out using Cell Quest (BD) and FlowJo (TreeStar)
software.
Peptide-dependent MHC-stabilization and -dissociation assays
These assays were performed, as reported (26), using test peptides at 1 mM.
RMA-S cells were primarily stained with the anti-Db/Ld mAb (HB-27) or
the anti-Kb mAb (HB-176; both from American Type Culture Collection).
Cells were secondarily stained with FITC-conjugated Rabbit Anti-Mouse
Ig (F0313; DakoCytomation). For MHC-dissociation assays, cells were
preincubated for 1 h at 37˚C with peptides, washed twice, reincubated
without peptides, and stained as described above. For a zero time point
determination, cells were pelleted immediately before (stabilization assay)
or after (dissociation assay) the incubation with peptide.
Analysis of viral transcripts
RNA was extracted from homogenized tumor samples stored in RNAlater
stabilization solution (Applied Biosystems) and from splenocytes following
the TRIzol protocol (Invitrogen). cDNA was synthesized using 5 mg RNA
with M-MuLV Reverse Transcriptase (Invitrogen). Quantitative PCR was
performed with an ABI 7900HT (Applied Biosystems) equipped with SDS
2.3 software. Primers and probes for viral transcripts were designed with
File Builder Software 3.1 and synthesized as Custom TaqMan Genomic
Assays (Applied Biosystems) using Gag forward primer, 59-CCGATCGTTTTGGACTCTTTGGT-39; Gag reverse primer, 59-TGTTTTAGGTTCTCGTCTCCTACCA-39; Gag probe, 59-CCCCTTAGAGGAGGGATAT-39;
Env forward primer, 59-ACTCAAGCTAGACCAGACAACTCAT-39; Env
reverse primer, 59-CCCCCACATGACTTGGATTCTC-39; and Env probe,
59-ATGAGGGATTTTATGTTTGCCC-39. Absolute quantification of Gag
and Env mRNAs was normalized to GAPDH mRNA (Mm99999915_g1;
Applied Biosystems) and expressed as copy number using a plasmid standard curve (18).
Adoptive-transfer experiments
Statistical analysis
Mice received total body gamma radiation (6 Gy) and were inoculated i.m.
with 150 ml the retroviral complex at day 0; the day after, mice received
a single i.v. administration of CTL clones (20 3 106 cells/mouse). For
homing experiments, equal numbers (10 3 106 cells/each) of Env- and
Gag-specific CTL clones were cotransferred in mice ∼12 d after radiation
and virus challenge. Tumor masses were processed 18–24 h later to recover infiltrating lymphocytes.
Mean values, SEM, Student t test, Mann–Whitney rank sum test, and
paired-samples Wilcoxon signed-rank test were calculated using MedCalc
software v11.3.1.0. A p value , 0.05 was considered statistically significant.
Cytolytic assays
In vitro cytotoxic activity was measured in a [51Cr]-release assay, as previously described (20), and data were calculated as the percentage of lysis
or as lytic units 30 (22). Functional avidity of CTLs was determined using
target cells pulsed with 3-fold dilutions of peptides (1025–10210 M for
Gag85–93 and 1028–10214 for Env189–196) and calculated as the peptide
concentration that resulted in 50% of maximal target cell–specific lysis
(23). In vivo cytotoxicity was measured as reported (24), with some
modifications. B6 splenocytes (107/ml) were labeled with three concentrations (5, 0.8, and 0.2 mM) of CFSE (Invitrogen), according to the
manufacturer’s indications. Thereafter, cells were pulsed with Gag85–93 or
Env189–196 peptides (1 mM) or left unpulsed. After washing, cell populations were mixed (10 3 106/each) and resuspended in PBS before i.v.
injection. Eighteen hours later, blood, spleen, and lymph node cell suspensions were analyzed with a FACSCalibur flow cytometer (Becton
Dickinson), and data were evaluated with Cell Quest software (Becton
Dickinson). Specific cytolytic activity was calculated as the difference
between the percentages of CFSEGag or CFSEEnv cells in control and virusinjected mice after normalization with CFSEunpulsed cell numbers.
Esterase activity assay
N-a-Benzyloxycarbonyl-L-lysin thiobenzyl–esterase secretion was evaluated as previously described (25). Gag- and Env-specific CTL clones (105)
were seeded in assay medium alone for spontaneous release or with 0.3 3
106 MBL-2 or with specific/control tetramers (1 mg/ml).
Cell staining and flow cytometry analysis
Stainings were carried out with the following: FITC-conjugated anti-CD44
mAb (clone IM7; BD Pharmingen), anti-CD49d mAb (VLA-4; clone R1-2;
BioLegend), anti-CD11a mAb (LFA-1; clone 2D7; BD Pharmingen) and
anti-CD8 mAb (CT-CD8a; Caltag); allophycocyanin-conjugated antiCD8a mAb (CT-CD8a; Caltag); PE-conjugated tetramers; CFSE (1 mM);
Results
IDD- and SDD-specific CTL clones disclose wide differences in
therapeutic efficacy and Ag avidity
Our previous characterization of T cell responses against the
transforming Mo-MSV/Mo-MuLV retroviral complex (18) demonstrated that virus-specific CTLs were primarily directed to the
Gag85–93/Db epitope, whereas a minority targeted the Env189–196/Kb
peptide, leading to the definition of these Ags as IDD and SDD,
respectively. To address the potential and relative therapeutic
efficacy of the two T cell subsets, we compared CTL clones recognizing either specificity in adoptive cell therapy (ACT) experiments (Fig. 1A). Daily monitoring revealed that only Gag-specific
CTL clones exerted complete protection against virus-induced
tumors, whereas clones directed to Env Ag were characterized
by reduced or no therapeutic efficacy.
To unveil the reasons for such differences, we investigated
phenotypic and functional properties of the different CTL
clones. CTL clones of either specificity efficiently recognized
and specifically killed target cells in cytotoxicity assays (Fig. 1B).
Degranulation capacity also was similar, as assessed by N-aBenzyloxycarbonyl-L-lysin thiobenzyl–esterase activity assays
(Supplemental Fig. 1A). Moreover, analysis of adhesion molecules potentially involved in T cell recirculation and homing
showed that the expression of CD44, VLA-4, and LFA-1 did not
vary substantially (Supplemental Fig. 1B). More importantly,
additional ACT experiments carried out in mice bearing advanced
tumors provided evidence that cotransferred CTL clones could be
retrieved in equal amounts at the tumor site (Gag/Env ratio, mean
1.11 6 1.19; Supplemental Fig. 1C). Conversely, peptide-titration
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
MBL-2, 293Db, and 293Kb cell lines (18) and the RMA-S cell line were
used. VAR-2 (H-2b) is an MBL-2 variant not expressing viral Ags that was
obtained in our laboratory.
4705
4706
FUNCTIONAL AVIDITY AND AICD SHAPE IMMUNODOMINANCE
experiments (Fig. 1C) disclosed that Env-specific clones recognized the Ag with an extremely high avidity (exceeding by 4 logs
that displayed by Gag-specific clones). Thus, a very high Ag
avidity is not intrinsically associated with CTL therapeutic efficacy upon adoptive transfer.
In vivo clonal dynamics of Gag- and Env-specific T cell
responses
The results reported above involved clonal populations that were
expanded in vitro and administered at the same amounts, which
likely do not reproduce the physiological development of antiviral
immune responses. Moreover, the higher TCR avidity of Envspecific CTLs apparently is in contrast with their previously reported subdominant role (18).
Therefore, we examined the generation kinetics of virus-specific
CTLs during the course of the immune response and tumor growth
(Supplemental Fig. 2A, 2B). After day 10, the percentage of Gagspecific TILs was higher than that of the Env-recognizing counterpart (Fig. 2A, Supplemental Fig. 2C). The difference between
Gag- and Env-specific T cells likely could not be ascribed to
a differential capacity to reach the tumor site, because they displayed an overlapping expression profile of the adhesion molecules that were tested (Supplemental Fig. 2D). These results
endorse the concept of ID of the Gag85–93 epitope, even though the
Env189–196-specific CTL population turned out to be more represented than expected from previous results (18).
A wide avidity difference characterizes Gag- and Env-specific
CTLs induced in vivo
Using an in vivo cytotoxicity assay (24), a striking lytic activity
was evident for the Gag specificity, being maximal when the
overall number of CTLs just started to decrease and the tumor was
shrinking (Fig. 2B). A relevant Env recognition also was detected,
even though it had a slightly retarded emergence, minor amplitude, and a faster disappearance. The delay in appearance of an
Env-specific cytotoxicity in vivo reflected the corresponding delay
in the amplification of the CTL subset found in tumor (Fig. 2A).
To study the effector abilities of TILs, tumors were collected at
the peak of the immune response. The strong activation of the
infiltrating population led to a cytotoxic activity that was detectable
directly ex vivo. Again, Gag-specific lytic activity was relatively
more pronounced than was that of the Env counterpart, at least
when the whole population was considered (Fig. 2C). However,
when considering the relative percentages of the two CTL specificities (Fig. 2A) and extrapolating the real E/T ratios, the Envspecific TILs were significantly more cytotoxic than the Gagspecific CTLs (Fig. 2D). Finally, when assayed for Ag avidity,
the Env-recognizing TIL population was at least four orders of
magnitude more avid (mean EC50 = 1.3 3 10211 6 6.1 3 10212 M)
than the Gag-specific TIL subset (mean EC50 = 2.3 3 1028 6 6.4 3
1029 M), fully reproducing what was seen with the CTL clones (twotailed p = 0.0037, Fig. 2E). Because of the higher avidity, Envspecific TILs turned out to be less CD8 dependent for their activity
compared with Gag-specific CTLs (Supplemental Fig. 2E).
T cell avidity is directly linked to a dose-dependent
proliferative block and apoptosis induction following Ag
in vitro restimulation
In vivo, CD8+ T cells are susceptible to proliferative inhibition by
high-dose peptide Ag, with an inverse correlation between the
concentration of Ag required for growth inhibition and CTL
functional avidity (27). Accordingly, in MLPCs, CTL expansion
was dose dependent for both Ags, even though the optimal concentration for Gag was 10–100-fold higher than that required for
Env (Fig. 3A), in line with the better MHC class I stabilization of
the latter (Supplemental Fig. 3). Nonetheless, even at the optimal
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 1. Adoptively transferred high-avidity T cell clones exert a reduced or no therapeutic efficacy against virus-induced tumors. (A) Therapeutic
activity of adoptively transferred CTL clones directed against immunodominant (Gag, clones 48 and 76) or subdominant (Env, clones 123 and 167) epitopes
in sublethally irradiated B6 mice injected with Mo-MSV/Mo-MuLV retroviral complex. As control, mice received naive syngeneic splenocytes or an anti–
H-2d allogeneic CTL clone. Number of mice tested is shown in parentheses; arrows indicate the day of CTL transfer. Data are mean 6 SD of virus-induced
tumor dimensions. (B) In vitro cytotoxicity of CTL clones. Lytic activity was evaluated against 51Cr-labeled Mo-MuLV–transformed MBL-2 cells and the
virus-negative VAR-2 variant loaded with viral or control peptides. The number of independent experiments carried out for each clone is shown in parentheses. (C) Functional avidity of Gag- and Env-specific CTL clones. 293Db and 293Kb cells pulsed with different concentrations of Gag85–93 and Env189–196
peptides, respectively, were used as targets in [51Cr]-release assays. The E/T ratio was fixed at 10:1. One of three independent experiments is shown.
The Journal of Immunology
4707
A higher avidity is also associated with a higher TCR expression,
described as cytofluorimetric brightness (15, 29), which, in turn, is
linked to a higher proneness to cell death. Accordingly, in all cell
cultures tested, both TCR brightness and annexin positivity of
Env-specific CTLs were significantly higher than the values
detected in the Gag population (Fig. 3D, 3E). Notably, a direct
correlation between TCR brightness and annexin positivity also
was evident within each Gag- or Env-specific T cell subpopulation
(Fig. 3F, 3G). Such behavior likely depended on activationinduced cell death (AICD) that was directly related to the level
of T cell avidity and peptide concentration.
Ag load and avidity of responding T cells influence in vivo
AICD susceptibility and lead to establishment of the
immunodominant and subdominant CTL responses
concentrations, Gag-specific CTL expansion overcame that detected
for the Env counterpart. In this regard, CFSE-labeled CTLs
restimulated with optimal Gag or Env doses showed that proliferation of these two antiviral CTL populations largely differed in size and timing (Fig. 3B).
Moreover, high-avidity CTLs have a greater sensitivity to Aginduced cell death (23, 27, 28). Accordingly, the Env-specific
CTL subset exhibited a significantly higher apoptotic index than
Gag-recognizing CTLs (Fig. 3C). In particular, the optimal Ag
dose was associated with the lowest annexin profile and with
maximal proliferation, whereas supraoptimal Ag concentrations
were associated with more pronounced apoptosis rates (Supplemental Fig. 4). Therefore, a combination of delayed growth
and high apoptosis index characterizes the high-avidity Env-specific
T cell population and hampers its expansion. In contrast, loweravidity Gag-specific CTLs are less affected by these phenomena
and, ultimately, can continue their growth.
Discussion
In this study, we challenge the current dogma that immunodominant T cells are endowed with high functional avidity, a feature
that is generally associated with a more favorable immune response
in terms of magnitude, frequency, and efficacy (1, 2, 15, 23, 30,
31). In particular, this study discloses a novel mechanism leading
to the establishment of ID in CTL immune responses.
The analysis was carried out using the Mo-MSV/Mo-MuLV
retroviral model. This viral complex induces sarcomas that undergo spontaneous regression as a result of the generation of
a strong CTL response recognizing viral Ags as tumor-associated
Ags. Previously, we (18) identified its apparent immunodominant and subdominant antigenic determinants in the Gag85–93 and
Env189–196 epitopes, respectively, based on the analytical tools
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 2. The in vivo subdominant antiviral CTL response is characterized by a higher functional avidity than that of the dominant counterpart. (A) Whole expansion kinetics of CD8+/tet+ T cells determined in
tumor, spleen, and tumor-draining lymph node. Data are mean 6 SD of
four independent experiments that were carried out using 5–10 pooled
mice/time point in each experiment. (B) In vivo cytofluorimetric cytotoxicity assay in blood, spleen and tumor-draining lymph node at different
time points after virus injection. Peptide-pulsed, CFSE-labeled syngeneic
splenocytes were inoculated i.v. and used as target cells. (C) Ex vivo lytic
activity of TILs isolated at the peak of the immune response against 51Crlabeled Mo-MuLV–negative VAR-2 cells loaded with specific or control
peptides. Data are mean 6 SD of four independent experiments carried out
with pools of TILs from 7–10 mice. (D) Cytotoxic activity of Ag-specific
TILs calculated according to the percentages of CD8+/tetramer-specific T
lymphocytes within the whole population. (E) Functional avidity of TILs.
VAR-2 cells pulsed with different concentrations of relevant peptides were
used as targets in [51Cr]-release assays. The E/T ratio was fixed at 400:1. A
representative experiment is shown; the gray bars illustrate mean 6 SD
from three independent tests. *p , 0.05, **p , 0.01, Mann–Whitney rank
sum test (A, B, and D), Student t test (E).
To analyze whether the link between T cell avidity and AICD also
was operative in vivo, we quantified viral Ag load in tumor and
spleen and verified the proneness to apoptosis in the responding
T cell populations. Detectable levels of viral mRNA were found
from days 3 to 20 (Fig. 4A, data not shown); tumor mass accounted
for the highest levels of viral transcripts of both specificities, even
though Env mRNA was much more abundant than the Gag
counterpart. The maximal peak of production of transcripts was
coincident with the appearance of the CD8+ CTL response, and
transcript clearance was inversely correlated with TIL expansion.
Levels of viral transcripts were measurable, but almost negligible,
in spleen compared with tumors. Moreover, a detectable difference between Gag and Env TCR brightness was observable
in TILs but not in splenocytes (Fig. 4B); specifically, Envrecognizing TILs showed a TCR brightness that was significantly higher than that of the Gag-specific subset (Fig. 4C). Next,
we assessed apoptosis in Gag- and Env-specific CTLs (Fig. 4D).
In tumor, Env-reactive T cells were more prone to death than was
the Gag counterpart on any day tested; the difference was more
pronounced earlier, at the peak of viral transcripts, and subsequently faded when transcripts were nearly undetectable. In
spleen, the differences in annexin positivity between the two CTL
subsets were not significant. Notably, in accordance with in vitro
results, the greater TCR brightness found in Env-specific TILs
directly correlated with a greater tendency to undergo apoptosis
(Fig. 4E).
Overall, the present data support a link among Ag load, TCR
brightness, and susceptibility to apoptosis of the responding T cells.
Moreover, these features are strictly correlated with the different
CTL Ag avidity. Therefore, the Env-specific CTL population
undergoes a continuous loss that compromises its expansion. In
contrast, these phenomena affect Gag-specific T cells to a minor
extent, leading to the expansion of this population and ultimately
being responsible for its apparent ID.
4708
FUNCTIONAL AVIDITY AND AICD SHAPE IMMUNODOMINANCE
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 3. In vitro, antiviral CTL proliferation and apoptosis are a function of the Ag dose and the T cell Ag avidity. In all experiments shown,
splenocytes were collected at the peak of the immune response and restimulated with different peptide concentrations in MLPCs. (A) In vitro expansion of
Gag- and Env-specific CTLs was monitored daily by flow cytometry with specific tetramers and anti-CD8 mAb. Gag-specific CTLs did not expand at
peptide concentrations , 0.1 mM, and the corresponding curves are not shown. Data are mean 6 SD of five independent experiments; each culture was set
up by pooling cells from at least five mice. (B) FlowJo proliferation analysis of CFSE-labeled, Gag- and Env-specific CTLs during in vitro restimulation
with the optimal peptide concentration. The red peak identifies undivided parental cells. The percentage of divided parental cells is reported in the upper
right corner of each panel. A representative of three independent experiments is shown. (C) Kinetics of annexin profiles on CD8+/tetramer-specific CTLs.
Each graph compares annexin profiles of Gag- and Env-specific CTLs cultured with their optimal peptide concentration (1 and 0.1 mM for Gag and Env
peptides, respectively). Numbers in the upper left corner indicate the geometric mean of annexin staining. A representative of three independent
experiments is shown; each culture was set up by pooling splenocytes from five to seven mice. (D) Comparison of TCR brightness between Gag- and Envspecific CTLs at day 5 of restimulation in culture. Symbols represent individual cultures set up with pooled splenocytes at different peptide concentrations
(10, 1, or 0.1 mM for Gag peptide; 1, 0.1, 0.01, or 0.001 mM for Env peptide); experiments were repeated three times. (E) Comparison of annexin positivity
between Gag- and Env-specific CTLs; culture conditions as in (D). (F) Correlation of geometric mean values between annexin staining and TCR expression
levels (high or low intensity) in Gag-specific CTLs. (G) The same analysis in Env-recognizing CTLs. Coupled dots in (F) and (G) refer to paired subsets
within single cultures set up as in (D) and (E). **p , 0.01, Mann–Whitney rank sum test (D and E), Wilcoxon test for paired samples (F and G).
available at that time: a limiting-dilution analysis of CTL response. The concept of ID and subdominance of Gag- or Envresponding CTLs was simply based on their relative numbers at
the end of a complex phase of tumor initiation, growth, and regression. This satisfied the current definition of ID as an immune
response centered on one or just a few antigenic epitopes and
organized in a hierarchic and highly reproducible manner. However, no information was available on how such a hierarchy could
evolve during the interaction between the immune system and the
pathogen/tumor.
The Journal of Immunology
4709
The most important observations that we made are the following.
On the Ag side, Gag transcript was much less represented than
the Env counterpart during the course of viral infection, and the
derived “immunodominant” Gag85–93 epitope stabilized its MHCrestriction element less efficiently than did the apparent “subdominant” Env189–196 peptide; such features appear exactly the
opposite from what is expected for a typical IDD. On the T cell
side, a difference in Gag- and Env-specific CTL precursors did not
account for the establishment of the ID hierarchy, because they
were equally represented in all immune districts analyzed, at least
when they started to be detectable; thereafter, Gag-specific CTLs
were more abundant than Env-specific T cells, resulting in their
“numerical ID.” Surprisingly, the so called “subdominant” Envspecific CD8+ T cells were endowed with a very high functional
avidity for the cognate Ag, exceeding by four logs that for the
immunodominant counterpart. In this regard, it was reported
previously that T cells with high functional avidity are particularly
sensitive to a supraoptimal antigenic stimulation, with the consequent induction of AICD (27). In agreement with these observations, we demonstrated in vitro that, in our model, the Ag load
also influenced the expansion rate and the AICD, based on the
functional avidity of responding T cells. These features are likely
exacerbated by the superior stability of Env–MHC complexes,
which results in higher TCR occupancy and, in turn, potentially
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 4. A higher Ag avidity in conjunction with Ag overload in vivo leads to a greater apoptosis rate in anti-Env TILs. (A) Relative expression of
viral transcripts was determined by real-time RT-PCR in tumor (left panel) and spleen (right panel) at different time points after viral complex injection.
GAPDH was used as the housekeeping gene to normalize the experimental variability. Data (mean 6 SD) are a representative of three independent
experiments carried out with five to seven mice for each time point. (B and C) Comparative analysis of TCR brightness between Gag- and Env-specific
CTLs during the clearance of virus infection. (B) Representative ex vivo tetramer-staining profiles of tetramer-specific CTLs from tumor and spleen.
Colored lines and numbers in the upper left corner of each panel refer to the geometric mean of Gag-specific (black) and Env-recognizing (red) CTLs. In
(C), summarized data refer to TILs obtained in two different kinetics experiments in which five to seven animals were pooled at each time point. Coupled
dots refer to paired Ag-specific CTLs isolated at the same time point (days 11–14) and belonging to the same pool. (D and E) Comparative analysis of
annexin staining between Ag-specific Gag- and Env-specific CTLs during the virus clearance. (D) Annexin profiles of the tetramer-specific CTLs analyzed
in (B). (E) Annexin geometric mean data of tetramer-specific TILs summarized from the two kinetics shown in (C). In (B) and (D), each sample is a pool of
cells from five to seven animals; data are representative of two independent experiments. Again, in (B) and (D), data acquisition in the two anatomical sites
requires different cytofluorimetric compensations that preclude a direct comparison of fluorescence intensity between them. *p , 0.05, **p , 0.01,
Wilcoxon test for paired samples.
4710
FUNCTIONAL AVIDITY AND AICD SHAPE IMMUNODOMINANCE
In conclusion, the reciprocal influence between the Ags and the
responding T cells, rather than their biochemical and functional
properties per se, defines the fate of T cells, their ID hierarchy, and,
ultimately, the efficacy of the immune response. For these reasons,
the characteristics of one universally efficient T cell cannot be
generalized; only a comprehensive knowledge of the entire system
can allow the identification of the best TCR and the best CTL
population to be used in immunotherapeutic approaches.
Acknowledgments
We thank Dr. M. Bellone, Dr. V. Coppola, Dr. F. Dazzi, and Dr. F. Ronchese
for critical reading of the manuscript.
Disclosures
The authors have no financial conflicts of interest.
References
1. Yewdell, J. W., and J. R. Bennink. 1999. Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. Annu. Rev.
Immunol. 17: 51–88.
2. Chen, W., and J. McCluskey. 2006. Immunodominance and immunodomination:
critical factors in developing effective CD8+ T-cell-based cancer vaccines. Adv.
Cancer Res. 95: 203–247.
3. Probst, H. C., K. Tschannen, A. Gallimore, M. Martinic, M. Basler, T. Dumrese,
E. Jones, and M. F. van den Broek. 2003. Immunodominance of an antiviral
cytotoxic T cell response is shaped by the kinetics of viral protein expression. J.
Immunol. 171: 5415–5422.
4. van der Most, R. G., K. Murali-Krishna, J. G. Lanier, E. J. Wherry,
M. T. Puglielli, J. N. Blattman, A. Sette, and R. Ahmed. 2003. Changing
immunodominance patterns in antiviral CD8 T-cell responses after loss of epitope presentation or chronic antigenic stimulation. Virology 315: 93–102.
5. Brett, S. J., K. B. Cease, and J. A. Berzofsky. 1988. Influences of antigen processing on the expression of the T cell repertoire. Evidence for MHC-specific
hindering structures on the products of processing. J. Exp. Med. 168: 357–373.
6. Tenzer, S., E. Wee, A. Burgevin, G. Stewart-Jones, L. Friis, K. Lamberth,
C. H. Chang, M. Harndahl, M. Weimershaus, J. Gerstoft, et al. 2009. Antigen
processing influences HIV-specific cytotoxic T lymphocyte immunodominance.
Nat. Immunol. 10: 636–646.
7. Rodriguez, F., S. Harkins, M. K. Slifka, and J. L. Whitton. 2002. Immunodominance in virus-induced CD8(+) T-cell responses is dramatically modified by
DNA immunization and is regulated by gamma interferon. J. Virol. 76: 4251–
4259.
8. Kedl, R. M., J. W. Kappler, and P. Marrack. 2003. Epitope dominance, competition and T cell affinity maturation. Curr. Opin. Immunol. 15: 120–127.
9. van der Most, R. G., A. Sette, C. Oseroff, J. Alexander, K. Murali-Krishna,
L. L. Lau, S. Southwood, J. Sidney, R. W. Chesnut, M. Matloubian, and
R. Ahmed. 1996. Analysis of cytotoxic T cell responses to dominant and subdominant epitopes during acute and chronic lymphocytic choriomeningitis virus
infection. J. Immunol. 157: 5543–5554.
10. Chen, W., S. Khilko, J. Fecondo, D. H. Margulies, and J. McCluskey. 1994.
Determinant selection of major histocompatibility complex class I-restricted
antigenic peptides is explained by class I-peptide affinity and is strongly influenced by nondominant anchor residues. J. Exp. Med. 180: 1471–1483.
11. Brennan, R. M., J. J. Miles, S. L. Silins, M. J. Bell, J. M. Burrows, and
S. R. Burrows. 2007. Predictable alphabeta T-cell receptor selection toward an
HLA-B*3501-restricted human cytomegalovirus epitope. J. Virol. 81: 7269–
7273.
12. Brawand, P., G. Biasi, C. Horvath, J. C. Cerottini, and H. R. MacDonald. 1998.
Flow-microfluorometric monitoring of oligoclonal CD8+ T cell responses to an
immunodominant Moloney leukemia virus-encoded epitope in vivo. J. Immunol.
160: 1659–1665.
13. Pion, S., G. J. Christianson, P. Fontaine, D. C. Roopenian, and C. Perreault.
1999. Shaping the repertoire of cytotoxic T-lymphocyte responses: explanation
for the immunodominance effect whereby cytotoxic T lymphocytes specific for
immunodominant antigens prevent recognition of nondominant antigens. Blood
93: 952–962.
14. La Gruta, N. L., W. T. Rothwell, T. Cukalac, N. G. Swan, S. A. Valkenburg,
K. Kedzierska, P. G. Thomas, P. C. Doherty, and S. J. Turner. 2010. Primary CTL
response magnitude in mice is determined by the extent of naive T cell recruitment and subsequent clonal expansion. J. Clin. Invest. 120: 1885–1894.
15. Alexander-Miller, M. A. 2005. High-avidity CD8+ T cells: optimal soldiers in
the war against viruses and tumors. Immunol. Res. 31: 13–24.
16. Viganò, S., D. T. Utzschneider, M. Perreau, G. Pantaleo, D. Zehn, and A. Harari.
2012. Functional avidity: a measure to predict the efficacy of effector T cells?
Clin. Dev. Immunol. 2012: 153863.
17. Chieco-Bianchi, L., D. Collavo, and G. Biasi. 1988. Immunologic unresponsiveness to murine leukemia virus antigens: mechanisms and role in tumor development. Adv. Cancer Res. 51: 277–306.
18. Milan, G., A. Zambon, M. Cavinato, P. Zanovello, A. Rosato, and D. Collavo.
1999. Dissecting the immune response to moloney murine sarcoma/leukemia
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
lethal activation levels in responding T cells (32). Indeed, the
highly avid Env-specific CTLs were affected by a constantly
higher apoptotic index that ultimately led to their impaired/failed
expansion. Therefore, this phenomenon resulted in the “apparent”
ID of the Gag-specific CTL subset. Nonetheless, the Ag load
primarily impacted the expansion of responding T cells, leaving
their effector capacity completely unaffected, as reported in other
models (27). This allows the exclusion of functional exhaustion as
a determinant in the establishment of the ID hierarchy, which is
different from what is observed during lymphocytic choriomeningitis virus (3) and HIV infection (16). Most importantly, the
role of Ag load, functional avidity, and AICD as a conceptual link
in the determinism of ID also was demonstrated clearly in vivo.
Indeed, we showed that the more avid Env-specific TILs face
a higher Ag load and, hence, undergo a more pronounced AICD,
allowing the final fictitious ID of Gag-specific T cells. Conversely,
in vivo, splenocytes of both specificities displayed a similar TCR
affinity/brightness and an equal susceptibility to cell death, because they encounter a similar Ag load that is considerably lower
than in the tumor. These data can shed some light on the therapeutic failure of Env-specific T cell clones in ACT. Indeed, the
impressive functional avidity of these effectors and the higher Ag
burden likely condemned Env-specific clones to a premature and
severe AICD, before they could perform any evident therapeutic
activity.
Overall, we demonstrated a physiological and elegant process
capable of shaping the immune response. Although the reciprocal
influence between the Ag dose and the functional avidity of
a responding T cell population is in line with other reports (33–35),
we propose the existence of an AICD-driven inverse correlation
between functional avidity and the establishment of the ID hierarchy. Therefore, this novel mechanism in the establishment of the
ID restrains and opposes the expansion of high-avidity T cells in
favor of lower-affinity populations.
Our results provide a mechanistic insight about several previously reported observations, in particular in infectious diseases but
also in cancer. Lichterfeld et al. (36) showed that, during the course
of HIV infection, the virus-responding higher-avidity CD8+ T
cells are selectively lost, leaving CTLs with lower functional avidity to become immunodominant. They proposed, but did not demonstrate, that the former CTLs should undergo AICD, particularly in
the presence of persistent high-level viremia. In contrast, Molldrem et al. (37) reported that, in chronic myelogenous leukemia,
the high tumor burden led to the selective deletion of high-avidity
T cells specific for a leukemic tumor Ag in a sort of reverse
immunoediting.
Therefore, the Ag load can selectively shape the overall functional avidity of a T cell population by expanding and/or deleting
distinct cell subsets that are characterized by the best-fitting or
higher functional avidity CTLs, respectively. In addition to the Ag
dose and presentation (33, 38), all of the factors impacting the fine
balance between Ag and responding T cells (the immunization
route (39), the biochemical nature of the immunogen [DNA versus peptide (40)], and possibly the persistence of the Ag (41)
acquire even more importance when dealing with high-avidity
T cells because of their proneness to AICD in an excess of
stimulation.
This last consideration implies that the selection of the most avid
CTLs for adoptive immunotherapy might not necessarily improve
the outcome of immunotherapeutic strategies, because their transfer in the presence of a high tumor burden could lead to a kind
of “immune kamikazing.” Conversely, T cells with lower functional avidity could potentially distinguish between tumor and
healthy tissues based on the differential Ag levels displayed (42).
The Journal of Immunology
19.
20.
21.
22.
23.
24.
26.
27.
28.
29.
30. Alexander-Miller, M. A., G. R. Leggatt, and J. A. Berzofsky. 1996. Selective
expansion of high- or low-avidity cytotoxic T lymphocytes and efficacy for
adoptive immunotherapy. Proc. Natl. Acad. Sci. USA 93: 4102–4107.
31. Kitazono, T., T. Okazaki, N. Araya, Y. Yamano, Y. Yamada, T. Nakamura,
Y. Tanaka, M. Inoue, and S. Ozaki. 2011. Advantage of higher-avidity CTL
specific for Tax against human T-lymphotropic virus-1 infected cells and tumors.
Cell. Immunol. 272: 11–17.
32. Valitutti, S., S. M€uller, M. Dessing, and A. Lanzavecchia. 1996. Different
responses are elicited in cytotoxic T lymphocytes by different levels of T cell
receptor occupancy. J. Exp. Med. 183: 1917–1921.
33. Luciani, F., M. T. Sanders, S. Oveissi, K. C. Pang, and W. Chen. 2013. Increasing
viral dose causes a reversal in CD8+ T cell immunodominance during primary
influenza infection due to differences in antigen presentation, T cell avidity, and
precursor numbers. J. Immunol. 190: 36–47.
34. Viganò, S., F. Bellutti Enders, I. Miconnet, C. Cellerai, A. L. Savoye, V. Rozot,
M. Perreau, M. Faouzi, K. Ohmiti, M. Cavassini, et al. 2013. Rapid perturbation
in viremia levels drives increases in functional avidity of HIV-specific CD8
T cells. PLoS Pathog. 9: e1003423.
35. Holbrook, B. C., R. D. Yammani, L. K. Blevins, and M. A. Alexander-Miller.
2013. In vivo modulation of avidity in highly sensitive CD8(+) effector T cells
following viral infection. Viral Immunol. 26: 302–313.
36. Lichterfeld, M., X. G. Yu, S. K. Mui, K. L. Williams, A. Trocha,
M. A. Brockman, R. L. Allgaier, M. T. Waring, T. Koibuchi, M. N. Johnston,
et al. 2007. Selective depletion of high-avidity human immunodeficiency virus
type 1 (HIV-1)-specific CD8+ T cells after early HIV-1 infection. J. Virol. 81:
4199–4214.
37. Molldrem, J. J., P. P. Lee, S. Kant, E. Wieder, W. Jiang, S. Lu, C. Wang, and
M. M. Davis. 2003. Chronic myelogenous leukemia shapes host immunity by
selective deletion of high-avidity leukemia-specific T cells. J. Clin. Invest. 111:
639–647.
38. Kim, M., H. B. Moon, K. Kim, and K. Y. Lee. 2006. Antigen dose governs the
shaping of CTL repertoires in vitro and in vivo. Int. Immunol. 18: 435–444.
39. Lin, L. C., I. E. Flesch, and D. C. Tscharke. 2013. Immunodomination during
peripheral vaccinia virus infection. PLoS Pathog. 9: e1003329.
40. Brentville, V. A., R. L. Metheringham, B. Gunn, and L. G. Durrant. 2012. High
avidity cytotoxic T lymphocytes can be selected into the memory pool but they
are exquisitely sensitive to functional impairment. PLoS ONE 7: e41112.
41. Hailemichael, Y., Z. Dai, N. Jaffarzad, Y. Ye, M. A. Medina, X. F. Huang,
S. M. Dorta-Estremera, N. R. Greeley, G. Nitti, W. Peng, et al. 2013. Persistent
antigen at vaccination sites induces tumor-specific CD8+ T cell sequestration,
dysfunction and deletion. Nat. Med. 19: 465–472.
42. Zhong, S., K. Malecek, L. A. Johnson, Z. Yu, E. Vega-Saenz de Miera,
F. Darvishian, K. McGary, K. Huang, J. Boyer, E. Corse, et al. 2013. T-cell
receptor affinity and avidity defines antitumor response and autoimmunity in
T-cell immunotherapy. Proc. Natl. Acad. Sci. USA 110: 6973–6978.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
25.
virus-induced tumors by means of a DNA vaccination approach. J. Virol. 73:
2280–2287.
Schepers, K., M. Toebes, G. Sotthewes, F. A. Vyth-Dreese, T. A. Dellemijn,
C. J. Melief, F. Ossendorp, and T. N. Schumacher. 2002. Differential kinetics of
antigen-specific CD4+ and CD8+ T cell responses in the regression of retrovirusinduced sarcomas. J. Immunol. 169: 3191–3199.
Rosato, A., S. Dalla Santa, A. Zoso, S. Giacomelli, G. Milan, B. Macino,
V. Tosello, P. Dellabona, P. L. Lollini, C. De Giovanni, and P. Zanovello. 2003.
The cytotoxic T-lymphocyte response against a poorly immunogenic mammary
adenocarcinoma is focused on a single immunodominant class I epitope derived
from the gp70 Env product of an endogenous retrovirus. Cancer Res. 63: 2158–
2163.
Rosato, A., S. Mandruzzato, V. Bronte, A. Zambon, B. Macino, F. Calderazzo,
P. Zanovello, and D. Collavo. 1995. Role of anti-LFA-1 and anti-ICAM-1
combined MAb treatment in the rejection of tumors induced by Moloney murine sarcoma virus (M-MSV). Int. J. Cancer 61: 355–362.
Rosato, A., A. Zoso, S. Dalla Santa, G. Milan, P. Del Bianco, G. L. De Salvo,
and P. Zanovello. 2006. Predicting tumor outcome following cancer vaccination
by monitoring quantitative and qualitative CD8+ T cell parameters. J. Immunol.
176: 1999–2006.
Derby, M. A., J. T. Snyder, R. Tse, M. A. Alexander-Miller, and J. A. Berzofsky.
2001. An abrupt and concordant initiation of apoptosis: antigen-dependent death
of CD8+ CTL. Eur. J. Immunol. 31: 2951–2959.
Camporeale, A., A. Boni, G. Iezzi, E. Degl’Innocenti, M. Grioni, A. Mondino,
and M. Bellone. 2003. Critical impact of the kinetics of dendritic cells activation
on the in vivo induction of tumor-specific T lymphocytes. Cancer Res. 63: 3688–
3694.
Zanovello, P., A. Rosato, V. Bronte, V. Cerundolo, S. Treves, F. Di Virgilio,
T. Pozzan, G. Biasi, and D. Collavo. 1989. Interaction of lymphokine-activated
killer cells with susceptible targets does not induce second messenger generation
and cytolytic granule exocytosis. J. Exp. Med. 170: 665–677.
Lipford, G. B., S. Bauer, H. Wagner, and K. Heeg. 1995. In vivo CTL induction
with point-substituted ovalbumin peptides: immunogenicity correlates with
peptide-induced MHC class I stability. Vaccine 13: 313–320.
Alexander-Miller, M. A., G. R. Leggatt, A. Sarin, and J. A. Berzofsky. 1996.
Role of antigen, CD8, and cytotoxic T lymphocyte (CTL) avidity in high dose
antigen induction of apoptosis of effector CTL. J. Exp. Med. 184: 485–492.
Derby, M., M. Alexander-Miller, R. Tse, and J. Berzofsky. 2001. High-avidity
CTL exploit two complementary mechanisms to provide better protection
against viral infection than low-avidity CTL. J. Immunol. 166: 1690–1697.
Laugel, B., H. A. van den Berg, E. Gostick, D. K. Cole, L. Wooldridge,
J. Boulter, A. Milicic, D. A. Price, and A. K. Sewell. 2007. Different T cell
receptor affinity thresholds and CD8 coreceptor dependence govern cytotoxic
T lymphocyte activation and tetramer binding properties. J. Biol. Chem. 282:
23799–23810.
4711