?E?7(CD103) Expression Identifies a Highly Active,
Tonsil-Resident Effector-Memory CTL Population1
Tonia Woodberry,2* Todd J. Suscovich,2* Leah M. Henry,* Meredith August,†
Michael T. Waring,* Amitinder Kaur,‡Christoph Hess,§Jeffery L. Kutok,¶Jon C. Aster,¶
Frederick Wang,?David T. Scadden,* and Christian Brander3*
The characterization of antiviral CTL responses has largely been limited to assessing Ag-specific immune responses in the pe-
ripheral blood. Consequently, there is an incomplete understanding of the cellular immune responses at mucosal sites where many
viruses enter and initially replicate and how the Ag specificity and activation status of CTL derived from these mucosal sites may
differ from that of blood-derived CTL. In this study, we show that EBV-specific CTL responses in the tonsils are of comparable
specificity and breadth but of a significantly higher magnitude compared with responses in the peripheral blood. EBV-specific,
tonsil-resident, but not PBMC-derived, T cells expressed the integrin/activation marker CD103 (?E?7), consistent with the de-
tection of its ligand, E-cadherin, on tonsillar squamous cells. These CD8-positive, CD103-positive, tonsil-derived CTL were largely
CCR7- and CD45RA- negative effector-memory cells and responded to lower Ag concentrations in in vitro assays than their
CD103-negative PBMC-derived counterparts. Thus, EBV-specific CTL in the tonsil, a crucial site for EBV entry and replication,
are of greater magnitude and phenotypically distinct from CTL in the peripheral blood and may be important for effective control
of this orally transmitted virus. The Journal of Immunology, 2005, 175: 4355–4362.
lifelong viral persistence (1, 2). Yet, relatively little is known about
virus-specific cellular immune responses at the site of viral entry.
In particular, the nature of antiviral lymphocytes in the tonsil, their
induction upon infection, and Ag specificity compared with lym-
phocytes in the peripheral blood is poorly understood. In the
present study, we compared EBV- and HIV-specific T cell re-
sponses in the tonsil and the peripheral blood to better understand
if, and how, virus-specific immune responses at the site of viral
entry and replication differ from those detected in the peripheral
EBV- and HIV-specific CD4 and CD8 T cell responses in the
peripheral blood have been well characterized during both acute
and chronic viral infection (3–8). Data from a number of studies
indicate that T cells from MALT can significantly differ in their
maturation status and phenotypes compared with peripheral blood-
derived T cells (9–17). Particularly, HIV- and CMV-specific ton-
sil- and blood-derived T cells have been shown to differ in their
perforin production, CD28 expression, and breadth of virus-spe-
cific CTL responses (17). A more recent, and more extensive,
pstein-Barr virus is one of the most prevalent human viral
infections, and the oral cavity has been identified as the
site of initial EBV infection as well as a reservoir for
study has also documented that although TCR repertoires of EBV-
specific CTL in the tonsil and PBMC were largely overlapping,
tonsil-derived cells expressed significantly more CD28 (15). To-
gether, these studies suggest that the local tonsil environment im-
parts functional and phenotypic differences on tonsil-resident CTL,
rather than merely reflecting the presence of different T cell pop-
ulations in various body compartments (15). Thus, potential func-
tional and phenotypic differences between peripheral and mucosal
T cells may be directly associated with the ability of lymphocytes
to circulate from the peripheral blood to mucosal sites, including
the intestinal mucosa and the tonsils.
Mucosal environments have indeed been shown to harbor lym-
phocytes that express a number of specific adhesion molecules,
including the intestinal homing receptors ?4?7and the integrin and
activation marker ?E?7(CD103) (18–23). In particular, ?4?7has
been shown to bind to the mucosal addressin mucosal addressin
cell adhesion molecule-1, which is selectively expressed on the
capillary endothelium in the gastrointestinal tract, whereas CD103
has been shown to bind E-cadherin, an integrin receptor known to
be expressed on intestinal endothelial cells, thymic tissue, and
hepatocytes (21, 24–26). Furthermore, CD103 has been described
as an intestinal intraepithelial lymphocyte and activation marker
(21, 22, 27). Thus, it is plausible that CD103 expression not only
allows homing or retention of T cells in these sites, but may also
be associated with an increased T cell activation status. Although
no study has demonstrated E-cadherin expression in the tonsil, the
finding that sputum-derived T lymphocytes and leukocytes in
bronchoalveolar lavages of healthy adults express significant
amounts of CD103 suggests that E-cadherin and CD103 could be
important for the retention of tonsil-resident CTL (22, 28–30).
Since the tonsil is a major reservoir for EBV, we assessed
whether EBV-specific cells isolated from tonsillar tissue would
differ in their phenotype, magnitude, and specificity compared with
cells isolated from the peripheral blood. Specifically, we asked
whether tonsillar T lymphocytes would express the integrin
CD103 and whether they would reveal a more activated effector
*Partners AIDS Research Center and†Oral and Maxillofacial Surgery, Massachusetts
General Hospital, Boston, MA 02129;‡New England Primate Research Center, Har-
vard Medical School, Southborough, MA 01772;§University Hospital, Basel, Swit-
zerland; and¶Department of Pathology and?Channing Laboratories, Brigham and
Women’s Hospital, Boston, MA 02115
Received for publication May 18, 2005. Accepted for publication July 28, 2005.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This study was supported by National Institutes of Health/National Institute for
Dental and Craniofacial Research Grant PO1 DE01438-01.
2T.W. and T.J.S. contributed equally to this work.
3Address correspondence and reprint requests to Dr. Christian Brander, Partners
AIDS Research Center, Fifth Floor, MGH East, No. 5214, 149 13th Street, Charles-
town, MA 02129. E-mail address: email@example.com
The Journal of Immunology
Copyright © 2005 by The American Association of Immunologists, Inc.0022-1767/05/$02.00
phenotype than their counterparts in the peripheral blood. This
enrichment in EBV- but not HIV-specific CTL in the tonsil, along
with the expression of the CD103 ligand, E-cadherin, on squamous
cells of the tonsillar epithelium provide a link between a viral
reservoir and the expansion of a site-specific immunity that may be
important for the control of viral replication and dissemination.
Materials and Methods
Study subjects and oral tissue samples
The characteristics of study subjects are summarized in Tables I and II.
Tonsil biopsies and peripheral blood samples were obtained from a total of
11 individuals, all with chronic EBV infection, as determined by the pres-
ence of detectable EBV-specific IgG (5). Three of these subjects were also
HIV coinfected and did (L859, L8154) or did not (K47) receive antiretro-
viral therapy. Oral samples were obtained from tonsillar biopsies in eight
individuals and from a sublingual biopsy in one subject. The tonsillar bi-
opsies yielded 6–100 million mononuclear cells, with no difference in the
yield from the HIV-negative or HIV-positive samples (median of 18 and 16
million, respectively). Two additional individuals donated parts of their
tonsils after a tonsillectomy performed to relieve sleep apnea. No inflam-
mation or other signs of tonsillitis were observed in any of the individuals.
The whole tonsils yielded 600 million and 1.1 billion mononuclear cells.
The study was approved by the Institutional Review Boards of Massachu-
setts General Hospital, and each subject provided written informed consent
before enrollment. For all subjects, peripheral blood samples were obtained
immediately before the biopsies were performed. PBMC were isolated by
density gradient centrifugation (Histopaque 1077; Sigma-Aldrich) within
24 h of venipuncture. Tonsil tissue was immediately placed on ice and
processed within 2 h of surgical removal by mechanical disruption using
scissors and passage over a 100-?m cell strainer (BD Labware). Isolated
tonsillar cells and PBMC were washed and tested directly without prior in
vitro expansion. High-resolution HLA typing was performed at the HLA
typing laboratory at Massachusetts General Hospital using sequence-spe-
cific primer PCR as previously described (31).
EBV-derived CTL epitopes used to screen for EBV-specific CTL re-
sponses have been previously described (5, 32). Optimal HIV class I-
restricted CTL epitopes were used to screen for HIV-specific cellular re-
sponses in the three HIV-coinfected subjects (33). Only described epitopes
known to be presented by the individual’s HLA class I type were used.
Peptides were synthesized at the peptide synthesis facility at Massachusetts
General Hospital using F-moc chemistry.
Tonsillar cells and PBMC were stained with CD3, CD4, CD8, and CD19
surface Abs (BD Biosciences) and analyzed by flow cytometry to deter-
mine the T and B cell ratios in each sample and to normalize the magnitude
of responses to the same input CD8 T cell number as indicated. Phenotypic
analyses were performed by staining cells with ?E?7(Immunotech)-, ?4?7
(BD Biosciences)-, CD69 (BD Biosciences)-, CXCR3 (BD Biosciences)-,
CCR7 (R&D Systems)-, CD45RA (BD Biosciences)-, or CD45RO (BD
Biosciences)-specific Abs as described elsewhere (34).
A total of 1 ? 106tonsillar cells or PBMC was incubated at 37°C for 30
min with tetramers and subsequently stained with surface Abs at room
temperature as described previously (35). Three different EBV-specific tet-
ramers were used, presenting either the HLA-B8-restricted lytic epitope
RAKFKQLL (B8-RAK) derived from EBV-BZLF1 protein, the epitope
FLRGRAYGL (B8-FLR) derived from EBNA-3A, or the HLA-A2-re-
stricted lytic epitope GLCTLVAML (A2-GLC) derived from EBV-
BMLF1 (5). Epitope-containing tetramers were prepared as described else-
Freshly isolated tonsillar cells and PBMC were tested directly in in vitro
ELISPOT assays as previously described (3). One hundred thousand to
200,000 PBMC and 100,000–300,000 tonsil cells were added to each well
in 100 ?l of RPMI 1640 supplemented with 10% FCS (Sigma-Aldrich).
Peptides were added at a final concentration of 10 ?g/ml. No peptide was
added to three to five wells, which served as negative controls. PHA (Re-
mel) was added at a concentration of 1.8 ?g/ml as a positive control. After
overnight incubation at 37°C, plates were developed and the number of
spots was determined using the AID ELISPOT Reader Unit (Autoimmun
Diagnostika). Results were expressed as spot-forming cells per 106input
CD8 T cells to compensate for variable CD8 T cell proportions in the tonsil
cells and the PBMC. Control experiments in which tonsil cells were sup-
plemented with peripheral blood cells or potent APC, such as EBV-trans-
formed B cells lines and PHA blasts did not show increased response rates
of tonsil cells, suggesting that APC function in the tonsil cell suspension
was not impaired. The cutoff for positive responses was determined as a
minimum of five spots per well or responses exceeding the mean of neg-
ative wells plus three times the SD, whichever gave the higher value.
Detection of E-cadherin in tonsil tissues
Immunohistochemistry was performed using 5-?m-thick Formalin-fixed,
paraffin-embedded tissue sections. Briefly, slides were soaked in xylene,
passed through graded alcohols, and put in distilled water. Slides were then
pretreated with 10 mM citrate, pH 6.0 (Zymed Laboratories) in a steam
pressure cooker (Decloaking Chamber; BioCare Medical) according to the
manufacturer’s instructions, followed by washing in distilled water. All
further steps were performed at room temperature in a hydrated chamber.
Slides were pretreated with Peroxidase Block (DakoCytomation) for 5 min
to quench endogenous peroxidase activity. Primary monoclonal anti-E-
cadherin Ab (DakoCytomation) was applied at a 1/400 dilution in Dako-
Cytomation diluent for 1 h. Slides were washed in 50 mM Tris-Cl (pH 7.4)
and anti-mouse HRP-conjugated Ab (Envision detection kit; DakoCyto-
mation) was applied for 30 min. After further washing, immunoperoxidase
staining was developed using a diaminobenzidine chromogen kit (Dako-
Cytomation) per the manufacturer’s protocol and counterstained with
Results are presented as median values unless otherwise noted. Statistical
analyses included the matched-pairs test (two tailed) for comparison of the
breadth, magnitude, and phenotype of immune responses detected in the
tonsil and the PBMC samples.
Increased magnitude of EBV latent Ag-specific T cells in tonsils
compared with PBMC
To assess virus-specific immune responses specific for an orally
transmitted virus (EBV) and a virus not typically transmitted orally
(HIV), tonsillar and peripheral blood samples were obtained from
a total of 11 EBV-infected subjects, 3 of whom were HIV-1 coin-
fected (Table I). The number of targeted CTL epitopes (breadth of
response) and the total magnitude of responses detected (total spot-
forming cells/106CD8 T cells) in the tonsil vs the peripheral blood
were assessed and compared in a direct ex vivo ELISPOT assay
using previously described, optimally defined EBV or HIV CTL
epitopes (32). Epitopes were selected based on the individual’s
HLA class I type, and between 7 and 32 different epitopes were
tested for each subject (5, 32, 37). The analyses revealed a com-
parable breadth of EBV-specific responses in tonsillar and periph-
eral T cells (Fig. 1A). Although a trend toward a greater breadth of
Table I. Cohort information
11.33 27.58 3.12 Tonsillar biopsy
2.28 13.44 5.6
2.32 7.27 2.7
66.68 15.38 3.12 Tonsillectomy
2.3 8.44 5.7
1.688.57 7.18 Tonsillar biopsy
7.40 3.15 Sublingual biopsy EBV
7.77.7 Tonsillar biopsy
Tonsillar biopsy EBV
4356CD103 EXPRESSION ON TONSIL-RESIDENT CTL
responses against latently expressed viral Ags in tonsillar T cells
compared with peripheral T cells was observed (p ? 0.052), the
number of lytic epitopes targeted in tonsillar and peripheral T cells
did not differ, as both compartments yielded a median of two de-
tected responses (data not shown).
In contrast, when the magnitude of the total EBV epitope-spe-
cific response was compared, significantly stronger responses were
detected in the tonsil compared with the blood (Fig. 1B; p ?
0.0064). This increase in magnitude was attributable to signifi-
cantly stronger responses to latent but not to lytic epitopes in the
tonsil (Fig. 1C; p ? 0.0026). Although the difference in total mag-
nitude could have been due to the slightly increased median num-
ber of responses detected in the tonsil, a comparison of the mag-
nitude of each individual response that was detectable in both
compartments showed a significantly increased magnitude in the
tonsil compared with the periphery (data not shown; p ? 0.0001).
These data were confirmed by intracellular cytokine analyses per-
formed in six individuals, again showing stronger latent responses
in tonsillar T cells compared with peripheral T cells (data not
shown). Together, the data demonstrate that the tonsils harbored
stronger EBV-specific T cell responses than the blood, with par-
ticularly strong tonsil-derived responses against epitopes derived
from viral proteins expressed during EBV latency.
Similar breadth and magnitude of HIV-specific CTL responses
in tonsillar cells and peripheral T cells
In contrast to EBV, the oral cavity is unlikely to serve as a major
reservoir for HIV and therefore HIV-specific CTL responses in the
tonsil may be of similar or reduced breadth and magnitude when
compared with the blood. To test this, cells from tonsil biopsies
obtained from three HIV-coinfected individuals were tested for
HIV epitope-specific immune responses and compared with re-
sponses in the peripheral blood. Among the 3 subjects, responses
to 12 HIV-derived CTL epitopes were tested in both compartments
(3, 4, and 5 epitopes per subject, respectively). A total of 11 pos-
itive responses were detected in PBMC, with 9 of them also in-
ducing positive responses in tonsillar tissue. Although there was
only a trend toward stronger HIV-specific responses in the blood
compared with the tonsils, EBV-specific responses were signifi-
cantly stronger in the tonsil compared with the blood for these
same three subjects (data not shown; p ? 0.008). Thus, HIV re-
sponses appeared equivalent in terms of their breadth and magni-
tude in the tonsil compared with the blood, whereas responses to
the orally transmitted EBV were increased in the tonsils of the
Tonsil-resident T cells express an effector-memory phenotype
and high levels of CD103
The increased frequency of EBV-specific T cells in the tonsillar
tissue may be due to the continual stimulation or immigration of
EBV-specific but not HIV-specific T cells in this compartment. To
test whether the presence of EBV-specific T cells in the tonsil was
indeed associated with the expression of specific adhesion or ac-
tivation markers, tonsil- and blood-derived T cells were analyzed
for the expression of adhesion/homing markers, including CXCR3,
?4?7and ?E?7(CD103), and markers discriminating effector and
naive T cells from effector-memory and central-memory T cells
Phenotyping showed a 2- to 10-fold higher percentage of CD8
T cells in the blood compared with the tonsils (p ? 0.001),
whereas CD4 T cells were enriched in tonsils relative to PBMC in
8 of the 10 individuals (Table II). The CD4 and CD8 T cells from
either compartment were tested for the expression of ?4?7, a mu-
cosal homing marker, and the chemokine receptor CXCR3, an-
other molecule involved in cell trafficking and expressed after re-
cent Ag contact (38, 43). The analyses revealed differences in the
expression of both markers, as CD8 T cells in the tonsil expressed
more frequently CXCR3 compared with the blood (Fig. 2A; p ?
0.034), whereas fewer tonsillar CD4 T cells expressed this homing
marker (p ? 0.027). In contrast, ?4?7expression was increased on
blood CD4 T cells compared with tonsillar CD4 cells (p ?
0.0001), yet its expression did not differ between CD8 T cells in
the two compartments (Fig. 2B). Tonsil-resident T cells were also
found to express ?E?7(CD103), an intraepithelial lymphocyte
marker expressed on mucosa-associated T lymphocytes as well as
on activated T cells and recent thymic emigrants (21, 25, 26, 44).
compared with PBMC. The breadth and magnitude of epitope-specific CTL
responses was determined in blood (Œ, ‚) and tonsil (F, E) samples ob-
tained either from HIV-negative (f, Œ, F) or HIV-infected subjects (?, ‚,
E). A, The total breadth of responses (number of epitopes targeted per
individual) is compared between the two compartments. B. The total mag-
nitude of responses (sum of all individual responses) is compared for each
subject between blood and tonsil. C, The total magnitude of responses
against epitopes encoded by latent (Œ and ‚ and F and E) or lytic (f and
? and ? and ?) EBV Ags in blood (Œ, ‚) and tonsil (F, E) is compared
for each subject. All responses are normalized to spot-forming cells per 106
input CD8 T cells.
Stronger EBV-specific responses are detected in tonsil cells
4357 The Journal of Immunology
On average, tonsil cells contained 16 times more CD103?CD8?T
cells and 3 times more CD103?CD4?T cells compared with
blood-derived CD4 and CD8 T cells (Fig. 2C and Table II; p ?
0.0001 for CD8). These data show that CD103 expression was
significantly elevated on tonsillar CD8 T cells compared with
PBMC, while ?4?7staining was comparable between tonsil and
blood-derived CD8 T cells.
The maturation and activation phenotype of T cells in the tonsil
and blood was further assessed using the previously described mat-
uration and activation markers CCR7 and the CD45RA/RO iso-
forms (40–42). Total lymphocytes, as well as EBV-specific CTL,
in tonsil samples or PBMC preparations, were stained with CCR7-
and CD45RA-specific Abs. Total lymphocyte staining was per-
formed on nine subjects, whereas for six subjects EBV-epitope
tetramers were available to assess CCR7/CD45 expression on
epitope-specific T cells in either compartment. A similar percent-
age of naive (CCR7?CD45RA?) and central memory cells
(CCR7?CD45RA?) were found in both the Ag-specific and non-
specific lymphocytes derived from the tonsil and the blood (data
not shown). In contrast, there were significant differences in the
percentages of effector-memory (CCR7?CD45RA?) and effector
(CCR7?CD45RA?)cells in both the total lymphocytes, as well as
the Ag-specific CD8 T cells between the two compartments. In
both cases, the tonsils expressed a higher percentage of effector-
memory cells than the blood (total lymphocytes, p ? 0.017 and Ag
specific, p ? 0.089). Conversely, effector cells were more fre-
quently detected in the blood than in the tonsils (total lymphocytes,
p ? 0.016 and Ag specific, p ? 0.043). Together, these data in-
dicate that the CD8 T cell population in the tonsil is enriched for
cells expressing an effector-memory phenotype both on total CD8
T cells and EBV Ag-specific CTL.
To show that epitope-specific cells with an effector-memory
phenotype also expressed CD103, tetramer analyses using EBV
latent and lytic Ag epitopes were combined with CD103 and
CD45RO staining (Fig. 3). The analyses included two individuals
tested against lytic protein-derived CTL epitopes and one individ-
ual tested for a latent Ag-specific response. Regardless of the or-
igin of the viral epitope, more tetramer-positive cells in the tonsil
were CD103?and CCR7?/CD45RO?compared with the periph-
eral blood. In all three cases, the Ag-specific, tetramer-positive
cells were at least 2-fold enriched in the tonsil compared with the
peripheral blood. In addition, CD103?effector-memory cell pop-
ulations were at least 15 times more prevalent in the tonsils com-
pared with those in blood. Similarly, CD103?CD45RO?cells
from the tonsil also more frequently expressed CD69 (65%) com-
pared with CD8?CD103?tonsillar T cells (26%; data not shown).
Together, the phenotypic analyses demonstrate that tonsil cells are
enriched for EBV-specific CTL and that these cells express spe-
cific markers of an activated, effector-memory phenotype.
CD103-expressing, tonsil-derived CTL are highly reactive to
low Ag concentration
Since CD103 has been described not only as an integrin but also as
an activation marker (45), we assessed whether the significant in-
crease in CD103?cells in the tonsil would be reflected by these
Table II. CD103 expression on T cells in tonsil tissue and PBMC
CD8 T Cells
CD4 T Cells
in tonsils and blood: CXCR3 (A), ?4?7(B), and CD103 (C) expression was
determined on CD8 (Œ, F) and CD4 (‚, E) T cells derived from blood (Œ,
‚) and tonsil (F, E). The percentage of cells expressing the respective
markers is indicated. Values of p are indicated only for comparisons where
statistically significant differences were observed.
Differential integrin/homing marker expression on T cells
4358CD103 EXPRESSION ON TONSIL-RESIDENT CTL
cells responding to decreasing Ag concentration. Total lympho-
cytes from the blood and the tonsil from two different individuals
were tested by ELISPOT assay using decreasing amounts of pep-
tide. As shown in Fig. 4, detectable responses were observed at 10-
to 100-fold lower concentrations of peptide in cells isolated from
the tonsil when compared with those from blood. These data were
confirmed in a third individual from whom tonsillar CD103?and
CD103?CD8 T cells were sorted by FACS and showed equal
magnitude of responses using 100 times lower peptide concentra-
tions in CD103?cells compared with CD103?cells (Fig. 4C).
Together, the data indicate that CD103 expression is associated
with a more activated, effector-memory-like phenotype in the ton-
sil, likely allowing these cells to respond to lower Ag concentra-
tions than CD103?cells.
The CD103 ligand E-cadherin is expressed in tonsils
CD103 has been identified on T cells derived from a number of
tissues, including the liver (24), renal and pancreatic allografts (46,
47), thymus (44, 48), and rectal and vaginal mucosa (12, 49). Al-
though alternative ligands for CD103 have been described (50), the
infiltration or, alternatively, retention of CD103-expressing cells in
the tonsil could be facilitated by the expression of E-cadherin in
this tissue. Indeed, histological analyses showed tonsillar squa-
mous epithelial cells strongly stained for E-cadherin, with no stain-
ing observed in infiltrating lymphoid cells (Fig. 5). This is the first
description of E-cadherin expression on tonsillar epithelial cells
and is consistent with the possibility that E-cadherin contributes to
the relative enrichment of CD103-expressing, EBV-specific CTL
in the tonsils.
The understanding of antiviral immune responses at mucosal sites
in the body is incomplete, and it is currently unclear whether the
peripheral blood and the oral mucosa are discrete immunological
compartments or whether there is immunological bidirectional
“spillover” between both sites. The present study aimed to identify
positive T cells was determined in the tonsil and blood by flow cytometry. Cells were stained with tetramers (A2-GLC (EBV BMLF-1 protein, GLCTL-
VAML), B8-RAK (EBV BZLF-1 protein, RAKFKQLL), and B8-FLR (EBV EBNA-3A protein, FLRGRAYG), as well as Abs against CD103 and
CD45RO, to indirectly assess the fraction of CD45RA-expressing, CD103?effector-memory cells.
Epitope-specific effector-memory cells from tonsil but not blood express CD103: the expression of CD103 on CD8-positive, tetramer-
spond to lower Ag concentrations more
effectively than peripheral T cells. Ton-
sil cells (–F–) and PBMC (- - -Œ- - -)
from two patients were analyzed in du-
plicate in an ELISPOT assay using se-
rial peptide dilutions of the BMRF-1-
YVLDHLIVV (A) or the EBNA-3c-de-
RRIYDLIEL (B). C, Tonsil cells from
subject K18 were separated by FACS
into CD103-positive and CD103-nega-
tive CD8 T cells. Forty thousand cells
per well were plated in an ELISPOT
assay and the LMP2-derived A2-CLG
peptide was added at concentrations
from 10 ?g/?l to 10 pg/?l. The graph
shows the number of spots detected at
the indicated peptide concentrations.
Insufficient CD8-positive CD103-posi-
tive tonsil T cells were obtained to test
the 1-ng/?l peptide concentration. All
responses are normalized to spot-form-
ing cells per 106input CD8 T cells.
Tonsillar T cells re-
4359 The Journal of Immunology
differences in the magnitude, the specificity, and the activation
status of antiviral CTL present between these sites and to as-
sociate these differences with the presence or absence of viral
replication in the different compartments. Surprisingly, initial
analyses revealed EBV-specific CTL responses of greater mag-
nitude in the tonsil compared with the peripheral blood, which
was largely mediated by increased CTL responses to epitopes
derived from latently expressed viral proteins. This increase of
EBV-specific CTL was observed even though the overall num-
ber of CD8 T cells in the tonsil was significantly reduced com-
pared with those of the peripheral blood. However, this reduc-
tion did not compromise the breadth of the EBV responses
The selective elevation of latent EBV-specific CTL responses
in the tonsil may reflect the persistent presence of B cells la-
tently infected with EBV. Although ongoing lytic Ag expres-
sion in the oral cavity has been described, it may be compara-
tively less frequent and possibly insufficient to maintain large
lytic protein-specific CTL populations (51). Alternatively, la-
tent EBV Ag-specific CTL responses in the peripheral blood
may be selectively reduced compared with lytic responses, po-
tentially due to reduced availability of latent Ag, since periph-
eral infected B cells express fewer EBV latent genes compared
with infected tonsillar B cells (52). Clearly, additional studies
in subjects with documented oral viral shedding at the time of
tonsil biopsy or tonsillectomy are needed to provide further
insight into the relationship between Ag expression and the
magnitude of EBV-specific cellular immunity in the oral cavity.
Furthermore, future studies would ideally include a more com-
prehensive testing of the total EBV-specific immunity. Because
the 90 EBV-derived CTL epitopes described to date (32) are
only derived from 12 different viral proteins, responses to other
regions may be missed. Whether this would affect the responses
detected in the tonsil, the blood, or both compartments will
require a complete knowledge of CTL epitopes, which may be
gained by screening tonsil in addition to PBMC against
Phenotypic analyses revealed significant differences between the
CD8 T cell populations in tonsils and blood. In particular, a sig-
nificantly higher proportion of the total and EBV-specific CD8 T
cell population in the tonsil expressed the E-cadherin-binding in-
tegrin ?E?7(CD103) when compared with blood-derived CD8 T
cells (Fig. 2). Our demonstration of E-cadherin expression in en-
dothelial cells in the tonsil suggests that it serves as the receptor for
CD103 on tonsil-resident cells (Fig. 5). However, CD103 also
binds the lymphocyte endothelial-epithelial cell adhesion molecule
for which expression throughout tonsillar epithelia as well as on
high endothelial venules in the tonsil has been demonstrated (50,
53). Thus, lymphoid cells in the crypt regions of tonsils, closely
associated with the surface epithelium lymphocytes that express
CD103, may bind either receptor, facilitating the immunosurveil-
lance of nonintestinal epithelia (54). In support of this, CD103
expression has also been documented on T cells in the lacrimal and
salivary glands of subjects with autoimmune diseases affecting the
oral cavity such as Sjo ¨gren’s syndrome (55). In addition, some
reports have found CD103 expression on what were considered
regulatory T cells, and a murine study suggests an important role
of CD103 in retaining these T cells at the site of Leishmania in-
fection (56, 57). However, in addition to CD103, other surface
markers known to play a role in lymphocyte homing and T cell
retention may also be important in populating the tonsil tissue with
virus-specific T cells. One possible candidate is CCR7, which is
more frequently expressed on CTL directed against EBV latent
Ags compared with CTL targeting lytic Ags, and which thus could,
at least in part, mediate the increased latent Ag-specific CTL ac-
tivity in the tonsil (16).
EBV-specific CD8 T cells and total CD8 T cells from tonsils
also expressed low levels of CD45RA and CCR7, consistent
with an effector-memory phenotype (40–42). In addition, these
largely CD103?, effector-memory tonsil CD8 T cells were more
reactive to their cognate Ag than the CD103?CD8 T cells (Fig.
4), since they required ?100 times lower Ag concentrations to
respond to Ag in vitro. These findings are in line with the pre-
viously reported role of CD103 as a marker of T cell activation
and suggests that retention of these highly sensitive cells in the
tonsil may permit rapid recall responses at even low Ag con-
centrations. Sample availability from oral biopsies did not per-
mit for the determination of EBV Ag, and larger numbers of full
tonsil donations would be needed to test for a potential associ-
ation between CTL responses viral gene expression in a statis-
tically satisfactory manner. Nevertheless, the present data sug-
gest a relationship between the anatomic location of viral
infection and the quality of the specific CTL response, provid-
ing a mechanism by which an orally transmitted virus may in-
duce different profiles of reactivity among lymphocytes in local
vs systemic compartments. If this relationship holds for other
viruses, it would have significant implications for vaccine
development and highlight the importance of a detailed
knowledge of cellular immune responses at specific sites of
pathogen exposure, rather than the limited focus on the
We thank Suqin He for generation of EBV tetramers.
The authors have no financial conflict of interest.
herin (diaminobenzidine chromogen, hematoxylin counterstain; original
magnification, ?200) show strongly positive squamous epithelium (A)
compared with negative staining control (B). Note that infiltrating lym-
phoid cells failed to stain for E-cadherin.
E-cadherin expression in tonsil. Tonsils stained for E-cad-
4360 CD103 EXPRESSION ON TONSIL-RESIDENT CTL
1. Sitki-Green, D., M. Covington, and N. Raab-Traub. 2003. Compartmentalization
and transmission of multiple Epstein-Barr virus strains in asymptomatic carriers.
J. Virol. 77: 1840–1847.
2. Walling, D. M., A. L. Brown, W. Etienne, W. A. Keitel, and P. D. Ling. 2003.
Multiple Epstein-Barr virus infections in healthy individuals. J. Virol. 77:
3. Frahm, N., B. T. Korber, C. M. Adams, J. J. Szinger, R. Draenert, M. M. Addo,
M. E. Feeney, K. Yusim, K. Sango, N. V. Brown, et al. 2004. Consistent cyto-
toxic-T-lymphocyte targeting of immunodominant regions in human immunode-
ficiency virus across multiple ethnicities. J. Virol. 78: 2187–2200.
4. Altfeld, M., E. S. Rosenberg, R. Shankarappa, J. S. Mukherjee, F. M. Hecht,
R. L. Eldridge, M. M. Addo, S. H. Poon, M. N. Phillips, G. K. Robbins, et al.
2001. Cellular immune responses and viral diversity in individuals treated during
acute and early HIV-1 infection. J. Exp. Med. 193: 169–180.
5. Woodberry, T., T. Suscovich, L. Henry, J. K. Davis, N. Frahm, B. Walker,
D. T. Scadden, F. Z. Wang, and C. Brander. 2005. Differential targeting and shifts
in the immunodominance of EBV specific CD8 and CD4 T cell responses from
acute to persistent infection. J. Infect. Dis. 192: 622–629.
6. Hislop, A. D., N. E. Annels, N. H. Gudgeon, A. M. Leese, and A. B. Rickinson.
2002. Epitope-specific evolution of human CD8?T cell responses from primary
to persistent phases of Epstein-Barr virus infection. J. Exp. Med. 195: 893–905.
7. Rickinson, A. B., and D. J. Moss. 1997. Human cytotoxic T lymphocyte re-
sponses to Epstein-Barr virus infection. Annu. Rev. Immunol. 15: 405–431.
8. Kaufmann, D. E., P. M. Bailey, J. Sidney, B. Wagner, P. J. Norris,
M. N. Johnston, L. A. Cosimi, M. M. Addo, M. Lichterfeld, M. Altfeld, et al.
2004. Comprehensive analysis of human immunodeficiency virus type 1-specific
CD4 responses reveals marked immunodominance of gag and nef and the pres-
ence of broadly recognized peptides. J. Virol. 78: 4463–4477.
9. Mowat, A. M., and J. L. Viney. 1997. The anatomical basis of intestinal immu-
nity. Immunol. Rev. 156: 145–166.
10. Masopust, D., V. Vezys, A. L. Marzo, and L. Lefrancois. 2001. Preferential
localization of effector memory cells in nonlymphoid tissue. Science 291:
11. Wherry, E. J., V. Teichgraber, T. C. Becker, D. Masopust, S. M. Kaech, R. Antia,
U. H. von Andrian, and R. Ahmed. 2003. Lineage relationship and protective
immunity of memory CD8 T cell subsets. Nat. Immunol. 4: 225–234.
12. Shacklett, B. L., T. J. Beadle, P. A. Pacheco, J. H. Grendell, P. A. Haslett,
A. S. King, G. S. Ogg, P. M. Basuk, and D. F. Nixon. 2000. Characterization of
HIV-1-specific cytotoxic T lymphocytes expressing the mucosal lymphocyte in-
tegrin CD103 in rectal and duodenal lymphoid tissue of HIV-1-infected subjects.
Virology 270: 317–327.
13. Shacklett, B. L., C. A. Cox, J. K. Sandberg, N. H. Stollman, M. A. Jacobson, and
D. F. Nixon. 2003. Trafficking of human immunodeficiency virus type 1-specific
CD8?T cells to gut-associated lymphoid tissue during chronic infection. J. Virol.
14. Shacklett, B. L., S. Cu-Uvin, T. J. Beadle, C. A. Pace, N. M. Fast, S. M. Donahue,
A. M. Caliendo, T. P. Flanigan, C. C. Carpenter, and D. F. Nixon. 2000. Quan-
tification of HIV-1-specific T-cell responses at the mucosal cervicovaginal sur-
face. AIDS 14: 1911–195.
15. Soares, M. V., F. J. Plunkett, C. S. Verbeke, J. E. Cook, J. M. Faint,
L. L. Belaramani, J. M. Fletcher, N. Hammerschmitt, M. Rustin, W. Bergler, et
al. 2004. Integration of apoptosis and telomere erosion in virus-specific CD8?T
cells from blood and tonsils during primary infection. Blood 103: 162–167.
16. Catalina, M. D., J. L. Sullivan, R. M. Brody, and K. Luzuriaga. 2002. Phenotypic
and functional heterogeneity of EBV epitope-specific CD8?T cells. J. Immunol.
17. Appay, V., P. Hansasuta, J. Sutton, R. D. Schrier, J. K. Wong, M. Furtado,
D. V. Havlir, S. M. Wolinsky, A. J. McMichael, D. D. Richman, et al. 2002.
Persistent HIV-1-specific cellular responses despite prolonged therapeutic viral
suppression. AIDS 16: 161–170.
18. Streeter, P. R., E. L. Berg, B. T. Rouse, R. F. Bargatze, and E. C. Butcher. 1988.
A tissue-specific endothelial cell molecule involved in lymphocyte homing. Na-
ture 331: 41–46.
19. Mora, A. L., and J. P. Tam. 1998. Controlled lipidation and encapsulation of
peptides as a useful approach to mucosal immunizations. J. Immunol. 161:
20. Cromwell, M. A., R. S. Veazey, J. D. Altman, K. G. Mansfield, R. Glickman,
T. M. Allen, D. I. Watkins, A. A. Lackner, and R. P. Johnson. 2000. Induction of
mucosal homing virus-specific CD8?T lymphocytes by attenuated simian im-
munodeficiency virus. J. Virol. 74: 8762–8766.
21. Cerf-Bensussan,N., A. Jarry,N.
D. Guy-Grand, and C. Griscelli. 1987. A monoclonal antibody (HML-1) defining
a novel membrane molecule present on human intestinal lymphocytes. Eur. J. Im-
munol. 17: 1279–1285.
22. Cepek, K. L., S. K. Shaw, C. M. Parker, G. J. Russell, J. S. Morrow, D. L. Rimm,
and M. B. Brenner. 1994. Adhesion between epithelial cells and T lymphocytes
mediated by E-cadherin and the ?E?7integrin. Nature 372: 190–193.
23. Berg, E. L., M. K. Robinson, R. A. Warnock, and E. C. Butcher. 1991. The
human peripheral lymph node vascular addressin is a ligand for LECAM-1, the
peripheral lymph node homing receptor. J. Cell Biol. 114: 343–349.
24. Shimizu, Y., M. Minemura, H. Murata, K. Hirano, Y. Nakayama, K. Higuchi,
A. Watanabe, T. Yasuyama, and K. Tsukada. 2003. Preferential accumulation of
CD103?T cells in human livers; its association with extrathymic T cells. J.
Hepatol. 39: 918–924.
25. Agace, W. W., J. M. Higgins, B. Sadasivan, M. B. Brenner, and C. M. Parker.
2000. T-lymphocyte-epithelial-cell interactions: integrin ?E(CD103)?7, LEEP-
CAM and chemokines. Curr. Opin. Cell Biol. 12: 563–568.
26. Taraszka, K. S., J. M. Higgins, K. Tan, D. A. Mandelbrot, J. H. Wang, and
M. B. Brenner. 2000. Molecular basis for leukocyte integrin ?E?7adhesion to
epithelial (E)-cadherin. J. Exp. Med. 191: 1555–1567.
27. Schieferdecker, H. L., R. Ullrich, A. N. Weiss-Breckwoldt, R. Schwarting,
H. Stein, E. O. Riecken, and M. Zeitz. 1990. The HML-1 antigen of intestinal
lymphocytes is an activation antigen. J. Immunol. 144: 2541–2549.
28. Leckie, M. J., G. R. Jenkins, J. Khan, S. J. Smith, C. Walker, P. J. Barnes, and
T. T. Hansel. 2003. Sputum T lymphocytes in asthma, COPD and healthy sub-
jects have the phenotype of activated intraepithelial T cells (CD69?CD103?).
Thorax 58: 23–29.
29. Erle, D. J., T. Brown, D. Christian, and R. Aris. 1994. Lung epithelial lining fluid
T cell subsets defined by distinct patterns of ?7and ?1integrin expression.
Am. J. Respir. Cell Mol. Biol. 10: 237–244.
30. Bernstein, J. M., R. Scheeren, E. Schoenfeld, and B. Albini. 1988. The distribu-
tion of immunocompetent cells in the compartments of the palatine tonsils in
bacterial and viral infections of the upper respiratory tract. Acta Otolaryngol.
Suppl. 454: 153–162.
31. Bunce, M., C. M. O’Neill, M. C. Barnardo, P. Krausa, M. J. Browning,
P. J. Morris, and K. I. Welsh. 1995. Phototyping: comprehensive DNA typing for
HLA-A, B, C, DRB1, DRB3, DRB4, DRB5 & DQB1 by PCR with 144 primer
mixes utilizing sequence- specific primers (PCR-SSP). Tissue Antigens 46:
32. Bihl, F. K., E. Loggi, J. V. Chisholm, H. S. Hewitt, L. M. Henry, C. Linde,
T. J. Suscovich, J. Wong, N. Frahm, P. Andreone, and C. Brander. 2005. Simul-
taneous assessment of cytotoxic T lymphocyte responses against multiple viral
infections by combined usage of optimal epitope matrices, anti-CD3 mAb T-cell
expansion and “RecycleSpot.” J. Translational Med. 3: 20–39.
33. Frahm, N., P. Goulder, and C. Brander. 2002. Total assessment of HIV specific
CTL responses: epitope clustering, processing preferences and the impact of HIV
C. B. B. Korber, B. Walker, R. Koup, J. Moore, B. Haynes, and G. Meyers, eds.
Los Alamos National Laboratory: Theoretical Biology and Biophysics, Los
34. Hess, C., T. K. Means, P. Autissier, T. Woodberry, M. Altfeld, M. M. Addo,
N. Frahm, C. Brander, B. D. Walker, and A. D. Luster. 2004. IL-8 responsiveness
defines a subset of CD8 T cells poised to kill. Blood 104: 3463–3471.
35. Goulder, P., Y. Tang, C. Brander, M. Betts, M. Altfeld, K. Annamalai, A. Trocha,
S. He, E. Rosenberg, G. Ogg, et al. 2000. Functionally inert HIV-specific cyto-
toxic T lymphocytes do not play a major role in chronically infected adults and
children. J. Exp. Med. 192: 1819–1832.
36. Goulder, P. J., M. A. Altfeld, E. S. Rosenberg, T. Nguyen, Y. Tang,
R. L. Eldridge, M. M. Addo, S. He, J. S. Mukherjee, M. N. Phillips, et al. 2001.
Substantial differences in specificity of HIV-specific cytotoxic T cells in acute
and chronic HIV infection. J. Exp. Med. 193: 181–194.
37. Frahm, N., P. Goulder, and C. Brander. 2003. Broad HIV-1 specific CTL re-
sponses reveal extensive HLA class I binding promiscuity of HIV-derived, op-
timally defined CTL epitopes. In HIV Molecular Immunology Database.
C. B. B. Korber, B. Walker, R. Koup, J. Moore, B. Haynes, and G. Meyers, eds.
Los Alamos National Laboratory: Theoretical Biology and Biophysics, Los
38. Sallusto, F., E. Kremmer, B. Palermo, A. Hoy, P. Ponath, S. Qin, R. Forster,
M. Lipp, and A. Lanzavecchia. 1999. Switch in chemokine receptor expression
upon TCR stimulation reveals novel homing potential for recently activated T
cells. Eur. J. Immunol. 29: 2037–2045.
39. Campbell, D., G. F. Debes, B. Johnston, E. Wilson, and E. C. Butcher. 2003.
Targeting T cell responses by selective chemokine receptor expression. Semin.
Immunol. 15: 277–286.
40. Hess, C., M. Altfeld, S. Y. Thomas, M. M. Addo, E. S. Rosenberg, T. M. Allen,
R. Draenert, R. L. Eldrige, J. van Lunzen, H. J. Stellbrink, et al. 2004. HIV-1
specific CD8?T cells with an effector phenotype and control of viral replication.
Lancet 363: 863–866.
41. Champagne, P., G. S. Ogg, A. S. King, C. Knabenhans, K. Ellefsen, M. Nobile,
V. Appay, G. P. Rizzardi, S. Fleury, M. Lipp, et al. 2001. Skewed maturation of
memory HIV-specific CD8 T lymphocytes. Nature 410: 106–111.
42. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two
subsets of memory T lymphocytes with distinct homing potentials and effector
functions. Nature 401: 708–712.
43. Kilshaw, P. J., and J. M. Higgins. 2002. ?E: no more rejection? J. Exp. Med. 196:
44. McFarland, R. D., D. C. Douek, R. A. Koup, and L. J. Picker. 2000. Identification
of a human recent thymic emigrant phenotype. Proc. Natl. Acad. Sci. USA 97:
45. Delgado, J., J. G. Bustos, M. C. Jimenez, E. Quevedo, and F. Hernandez-Navarro.
2002. Are activation markers (CD25, CD38 and CD103) predictive of sensitivity
to purine analogues in patients with T-cell prolymphocytic leukemia and other
lymphoproliferative disorders? Leuk. Lymphoma 43: 2331–2334.
46. Wang, D., R. Yuan, Y. Feng, R. El-Asady, D. L. Farber, R. E. Gress, P. J. Lucas,
and G. A. Hadley. 2004. Regulation of CD103 expression by CD8?T cells
responding to renal allografts. J. Immunol. 172: 214–221.
47. Feng, Y., D. Wang, R. Yuan, C. M. Parker, D. L. Farber, and G. A. Hadley. 2002.
CD103 expression is required for destruction of pancreatic islet allografts by
CD8?T cells. J. Exp. Med. 196: 877–886.
4361The Journal of Immunology
48. Kutlesa, S., J. T. Wessels, A. Speiser, I. Steiert, C. A. Muller, and G. Klein. 2002. Download full-text
E-cadherin-mediated interactions of thymic epithelial cells with CD103?
thymocytes lead to enhanced thymocyte cell proliferation. J. Cell Sci. 115:
49. Kaul, R., P. Thottingal, J. Kimani, P. Kiama, C. W. Waigwa, J. J. Bwayo,
F. A. Plummer, and S. L. Rowland-Jones. 2003. Quantitative ex vivo analysis of
functional virus-specific CD8 T lymphocytes in the blood and genital tract of
HIV-infected women. AIDS 17: 1139–1144.
50. Strauch, U. G., R. C. Mueller, X. Y. Li, M. Cernadas, J. M. Higgins,
D. G. Binion, and C. M. Parker. 2001. Integrin ?E(CD103)?7mediates adhesion
to intestinal microvascular endothelial cell lines via an E-cadherin-independent
interaction. J. Immunol. 166: 3506–3514.
51. Babcock, G. J., L. L. Decker, M. Volk, and D. A. Thorley-Lawson. 1998. EBV
persistence in memory B cells in vivo. Immunity 9: 395–404.
52. Babcock, G. J., D. Hochberg, and A. D. Thorley-Lawson. 2000. The expression
pattern of Epstein-Barr virus latent genes in vivo is dependent upon the differ-
entiation stage of the infected B cell. Immunity 13: 497–506.
53. Shieh, C. C., B. K. Sadasivan, G. J. Russell, M. P. Schon, C. M. Parker, and
M. B. Brenner. 1999. Lymphocyte adhesion to epithelia and endothelia mediated
by the lymphocyte endothelial-epithelial cell adhesion molecule glycoprotein.
J. Immunol. 163: 1592–1601.
54. Bernstein, D. C., and G. M. Shearer. 1988. Suppression of human cytotoxic T
lymphocyte responses by adherent peripheral blood leukocytes. Ann. NY Acad.
Sci. 532: 207–213.
55. Fujihara, T., H. Fujita, K. Tsubota, K. Saito, K. Tsuzaka, T. Abe, and
T. Takeuchi. 1999. Preferential localization of CD8??E?7?T cells around aci-
nar epithelial cells with apoptosis in patients with Sjo ¨gren’s syndrome. J. Immu-
nol. 163: 2226–2235.
56. Suffia, I., S. K. Reckling, G. Salay, and Y. Belkaid. 2005. A role for CD103 in
the retention of CD4?CD25?Treg and control of Leishmania major infection.
J. Immunol. 174: 5444–5455.
57. Rao, P. E., A. L. Petrone, and P. D. Ponath. 2005. Differentiation and expansion
of T cells with regulatory function from human peripheral lymphocytes by stim-
ulation in the presence of TGF-?. J. Immunol. 174: 1446–1455.
4362 CD103 EXPRESSION ON TONSIL-RESIDENT CTL