The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

The Journal of Immunology, 2007, 179, 4272 -4282
Copyright © 2007 by The American Association of Immunologists, Inc.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Winter, D.
Right arrow Articles by Maurer, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Winter, D.
Right arrow Articles by Maurer, D.

Down-Modulation of CXCR3 Surface Expression and Function in CD8+ T Cells from Cutaneous T Cell Lymphoma Patients1

Dorian Winter*,{dagger}, Julia Moser{dagger}, Ernst Kriehuber*,{dagger}, Christoph Wiesner{dagger}, Robert Knobler{ddagger}, Franz Trautinger{ddagger}, Paula Bombosi, Georg Stingl{dagger}, Peter Petzelbauer§, Antal Rot and Dieter Maurer2,*

* Research Center for Molecular Medicine (CeMM) of the Austrian Academy of Sciences, Vienna, Austria; {dagger} Division of Immunology, Allergy, and Infectious Diseases, {ddagger} Division of Special and Environmental Dermatology, and § Division of General Dermatology, Department of Dermatology, Medical University of Vienna, Vienna, Austria; and Novartis Institutes for Biomedical Research, Vienna, Austria


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Viruses can escape destruction by the immune system by exploitation of the chemokine-chemokine receptor system. It is less established whether human cancers can adopt similar strategies to evade immunologic control. In this study, we show that advanced cutaneous T cell lymphoma (CTCL) is associated with selective and efficient inactivation of CXCR3-dependent T cell migration. Our studies demonstrate that this alteration is at least in part due to CXCR3 down-regulation in vivo by elevated serum levels of CXCR3 ligands. The T cell population most affected by this down-regulatory mechanism are CD8+ cytotoxic effector T cells. In CTCL patients, cytotoxic effector T cells have strongly reduced surface CXCR3 expression, accumulate in peripheral blood, but are virtually absent from CTCL tumor lesions, indicating an inability to extravasate into lymphoma tissue. CTCL-associated inactivation of effector cell recruitment may be a paradigmatic example of a new type of immune escape mechanisms shielding the neoplasm from a tumoricidal attack.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Cutaneous T cell lymphoma (CTCL)3 is the most prevalent malignant lymphoproliferative disorder of the skin. Mycosis fungoides represents the vast majority of all CTCLs. In this disease, the transformed cell is a clonal CD4+CD7–CD26– memory T cell with a bizarrely shaped nucleus (1, 2, 3, 4). In early disease stage, these CTCL cells show a high degree of tropism for skin epithelium, whereas in later disease stages dermal infiltration predominates. Accordingly, scaly patches and plaques define early limited disease, whereas the occurrence of tumors indicates disease progression. Further disease progression is most often associated with hematogenic spreading of the tumor cells giving rise to T cell leukemia. The prognosis of mycosis fungoides patients is clearly linked to the extent of skin and, more importantly, systemic involvement. Cytodestructive and immunomodulatory therapies may induce partial remissions, but do not improve substantially the prognosis of advanced mycosis fungoides (5).

Recent studies indicate that CD8+ CTL can exert antitumor responses in mycosis fungoides, and that a down-regulation of CD8+ T cell function is associated with disease progression. Indeed, CD8+ T lymphocytes with specific cytotoxicity against parental tumor cells have been isolated from blood of mycosis fungoides patients. Candidate epitopes recognized by these lymphoma-specific CTL include variable and hypervariable regions of the clonal TCR and cancer-related Ags of the tumor/testis family (6, 7, 8, 9). Interestingly, lymphoma-specific cytotoxic T cell responses of comparable specificity and magnitude could be induced from the blood of healthy controls and of lymphoma patients with a far advanced disease (6). This indicates that deletion or permanent anergy of lymphoma-specific circulating CD8+ T cells is not a viable explanation for a loss of immunologic control resulting in disease progression.

A relationship exists between a high percentage of cytotoxic CD8+ T lymphocytes in lymphoma skin infiltrates and a longer survival of patients (10, 11). This confirms that functional CD8+ CTLs have to enter lymphoma skin lesions to limit disease progression. Hence, an impairment of effector T cell homing into lymphoma lesions may underlie progression of CTCL from limited to advanced disease. Such condition could also explain why during disease progression from the plaque to the tumor stage, lesional cytotoxic effector CD8+ T cells are greatly reduced in numbers (10).

In this study, we investigate the possibility that CTCL impairs differentiation, cytotoxic function, and homing properties of CD8+ effector T cells. The patients studied in this work suffered from mycosis fungoides (subsequently referred to as CTCL) with or without hematologic tumor cell dissemination. We analyzed the expression and function of chemokine receptors on T cells of patients with advanced CTCL and found a selective defect in functional expression of CXCR3. In vitro studies suggest that this alteration is due to down-regulation of CXCR3 surface expression in vivo by elevated serum levels of cognate CXCR3 ligands (CXCR3L). These findings highlight a new chemokine receptor-based mechanism of hematologic malignancies to manipulate effector T cell migration and, thus, evade destruction. Consequently, therapeutic prevention of cancer-associated inhibition of effector T cell homing may be instrumental for the design of clinically effective cancer vaccines.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Patients

Heparinized blood (n = 25), serum, and 4- to 6-mm punch biopsies from lesional skin (n = 3) were obtained from mycosis fungoides (subsequently referred to as CTCL) patients with or without hematologic involvement (56.3% of patients had CD4+CD7– tumor cells in blood with a range from 17.8 to 88.9% (mean 51.3%) of all peripheral blood (PB) T cells). The number of patient samples analyzed in the various assays is given in the respective figure legends. The diagnosis of mycosis fungoides was based on clinical criteria, typical histology, and immunohistology in accordance with the classification of cutaneous lymphomas by the European Organization for Research and Treatment of Cancer (12). T cell clonality in skin and/or blood was determined by a standard multiplex PCR approach and/or by mAb-based anti-TCRVbeta immunophenotyping (6). Blood and lesional skin were obtained from one patient with varicella zoster virus (VZV) infection. Control blood samples were obtained from healthy donors, patients suffering from systemic scleroderma (undergoing cycles of extracorporeal photochemotherapy (ECP) every 4 wk), or patients after surgery for primary melanoma and treatment with s.c. administered IFN-{alpha} (three times per week). Blood samples from ECP- and/or IFN-{alpha}-treated CTCL and control patients were taken upon informed consent at the end of the interval between consecutive treatments.

Cells

PBMCs were prepared from heparinized blood by sedimentation over Ficoll-PaquePlus (Pharmacia). Remaining erythrocytes were lysed osmotically in a buffer containing 0.15 M NH4Cl, 1 mM KHCO3, and 0.1 mM EDTA (all from Sigma-Aldrich). CD8+ T cells were enriched by anti-CD4, anti-CD11b, anti-CD16, anti-CD19, anti-CD34, and anti-CD56 mAb (all by Beckman Coulter)-based immunomagnetic depletion (MACS; Miltenyi Biotec). By this procedure, CD8+ T cells were routinely enriched to a purity of >95%.

Cells from CTCL tumor lesions and VZV-induced skin lesions were isolated, as described previously (6). Briefly, punch biopsies were minced and cells were squeezed out by applying gentle manual pressure. Debris and fibers were removed by filtration through a 40-µm cell strainer (BD Discovery Labware Europe). The human erythroleukemia cell line K-562 was provided by O. Majdic, Institute of Immunology (Medical University of Vienna, Vienna, Austria).

Flow cytometry analysis

T cell populations were characterized using anti-CD3 (SK7) PerCP, anti-CD4 (Leu-3) PE, and anti-CD7 (4H9) FITC or anti-CD8 (SK1) FITC (all by BD Pharmingen). Anti-CD25 (2A3) PE, anti-HLA-DR (L243) PE, and anti-CD71 (M-A712) FITC were used to determine the activation status of CD8+ T cells. These analyses were performed on a FACScan cytometer (BD Biosciences). Anti-CCR1 (53504.111) biotin, anti-CCR2 (48607) PE, anti-CCR6 (53103.111) PE, anti-CCR8 (191704) PE, anti-CCR9 (112509) PE, anti-CXCR3 (49801.111) FITC, anti-CXCR5 (51505.111) PE, anti-CXCR6 (56811) PE (all from R&D Systems), anti-CCR4 (1G1) PE, anti-CCR5 (2D7) FITC, anti-CCR7 (3D12) biotin, anti-CXCR1 (5A12) PE, anti-CXCR2 (6C6) PE, and anti-CXCR4 (12G5) PE (all from BD Pharmingen) were used to analyze chemokine receptor expression of T cells. Cells were counterstained with anti-CD7 (4H9) FITC or PE, anti-CD8 (SK1) PerCP, and anti-CD3 (SK7) allophycocyanin (all from BD Pharmingen), and analyzed on a FACSCalibur flow cytometer (BD Biosciences). Subpopulations of CD8+ T cells were defined and characterized by stainings with anti-CD3 (UCHT-1) PC7 (Beckman Coulter), anti-CD8 PerCP, anti-CD27 (L128) FITC, and anti-CD45RA (HI100) allophycocyanin (both from BD Pharmingen). In selected experiments, cells were counterstained further by anti-CCR5, anti-CCR7, or anti-CXCR3. Five-color analyses and cell sorting were performed on a FACSAria flow cytometer (BD Biosciences).

For intracellular stainings, PBMC, either freshly isolated or incubated for 5 days with 100 U/ml IL-2 (Strathmann Biotech), were washed and resuspended in PBS at a density of 1 x 106 cells/ml. Aliquots of 1 x 105 cells were stained with anti-CD3 (SK7) allophycocyanin and anti-CD8 (SK1) PE (both from BD Pharmingen). After two washings, cells were fixed and permeabilized using the FIX&PERM reagent (An der Grub Bio Research), according to the instructions of the manufacturer. Anti-CXCR3 FITC or appropriate isotype-matched control mAbs were used for intracellular staining. Cells were analyzed by flow cytometry on a FACSCalibur.

CD3-dependent killing assay

Fc{gamma}R-positive K-562 cells were labeled with 0.1 µM CFSE (Molecular Probes). CFSE-labeled K-562 cells were incubated in PBS for 1 h at 4°C in the presence or absence of 50 µg/ml anti-CD3 mAb (OKT-3; DakoCytomation). After two washes in ice-cold PBS, 2 x 105 cells/well were plated in 96-well U-bottom plates (Corning Glass), and purified CD8+ T cells were added. Target cells alone were plated for control purposes. After gentle spinning at 150 x g for 5 min, cells were cultured for 4 h at 37°C. Then cells were collected and incubated with 10 µM 7-aminoactinomycin D (7-AAD; Sigma-Aldrich). After two washes, cells were analyzed by flow cytometry, and the ratio of 7-AAD+ to 7-AAD– CFSE+ cells was determined (13). Effector T cell-mediated, CD3-dependent target cell killing was expressed as the mean percentage of 7-AAD+ target cells in the presence of OKT3 minus the mean percentage of 7-AAD+ target cells in its absence. The assays were performed in triplicate.

CD3-dependent redirected lysis

K-562 cells were labeled with Eu3+, as described previously (6, 14). Briefly, 5 x 106 cells were incubated on ice for 15 min in 1 ml of labeling buffer (0.125 mM EuCl3 (Fluka), 0.625 mM N, N-bis[2-(bis[carboxymethyl]amino)-ethyl]glycine (Sigma-Aldrich)). After addition of 20 µl of 0.1 M CaCl2 (Merck), cells were washed three times in a buffer containing 10 mM D+ glucose (Sigma-Aldrich) and 2 mM CaCl2. Labeled K-562 were exposed to OKT-3 (50 µg/ml) for 1 h in serum-free IMDM. After two washes in ice-cold PBS, viable cells were isolated on 50% Percoll gradients (Pharmacia). A total of 2 x 103 cells/well was plated into 96-well U-bottom plates (Corning Glass) as targets for purified CD8+ T cells. Target cells alone were plated to determine spontaneous Eu3+ release. After gentle spinning at 150 x g for 5 min, cells were incubated for 4 h at 37°C. A total of 20 µl of the supernatant was mixed with 200 µl of Enhancement Solution (Wallac), and Eu3+ release was quantified by measuring time-resolved fluorescence on a 1234 DELFIA fluorometer (Wallac). Specific target cell lysis was calculated as follows: (experimental release – spontaneous release)/(maximum release – spontaneous release) x 100. Maximum release was induced by incubation with 2% Triton X-100 (Sigma-Aldrich). The assays were performed in triplicate.

Confocal microscopy

Purified CD8+ T cells were resuspended at a density of 2 x 106 cells/ml in PBS. Aliquots containing 4 x 104 cells were bound to adhesion slides (Bio-Rad) for 30 min at 4°C. Adherent cells were fixed and permeabilized using FIX&PERM. During permeabilization, cells were incubated with anti-CXCR3 (R&D Systems; diluted 1/20) or appropriate isotype control mAbs and rabbit anti-human CD3 (OKT-3; diluted 1/10; DakoCytomation). After washings with cold PBS/1% BSA, cells were incubated with Fab anti-mouse Alexa 488 (1:200; Molecular Probes) and Fab anti-rabbit Cy5 (1:200; Jackson ImmunoResearch Laboratories). Slides were mounted in Fluoroprep (bioMerieux). To define the intracellular localization of CXCR3, purified CD8+ T cells were fixed in ice-cold acetone for 5 min, blocked in PBS/5% BSA/5% goat serum, and exposed simultaneously to FITC-conjugated anti-lysosome-associated membrane protein (LAMP)2 mAbs (1:100), PE-conjugated anti-CXCR3 mAbs (1:100; both from BD Pharmingen), and rabbit anti-early endosomal Ag (EEA)1 (1:500; Abcam) at 4°C overnight. After several washes in blocking solution, allophycocyanin-conjugated goat anti-rabbit IgG (1:250; Invitrogen Life Technologies) was applied for 45 min. Slides were mounted in AquaTex (Merck). Cryosections (7 µm) of CTCL skin lesions were fixed in acetone/4% formaldehyde. After permeabilization in 0.5% Triton X-100/PBS for 5 min, the sections were incubated for 45 min with mAbs against CXCL11 or CXCL9 (R&D Systems) and mAb binding was revealed by Alexa 488- or Alexa 546-conjugated secondary Abs (Invitrogen Life Technologies). In double-labeling experiments, free binding sites of bound secondary Abs were saturated by irrelevant mouse IgG before exposure to FITC-conjugated anti-HLA-DR mAbs (BD Pharmingen). Sections were mounted in Vectashield Hard Set (Vector Laboratories). Cells and sections were examined using a LSM 520 confocal microscope (Zeiss).

Real-time quantitative PCR

CD8+ T cells were isolated from CTCL patients and healthy donors and divided in aliquots of 1 x 106 cells before or after 5-day incubation with 100 U/ml IL-2. mRNA was extracted (mRNA isolation kit; Roche Diagnostics) and reverse transcribed (first strand cDNA synthesis kit for RT-PCR; Roche). TaqMan RT-PCR (Applied Biosystems) was performed with the following primers and probes: CXCR3A, FAM probe 5'-TGAGTGACCACCAAGTGCTAAATGACGC-3', forward primer 5'-ACCCAGCAG CCAGAGCACC-3', and reverse primer 5'-TCGGCGTCATTTAGC; ACTTG-3', CXCR3B, FAM probe 5'-CCCGTTCCCGCCCTCACAGG-3', forward primer 5'-TGCCAGGCCTTTACACAGC-3', and reverse primer 5'-TCGGCGTCATTTAGCACTTG-3' (15). beta2-microglobulin (beta2m) quantification for normalization was performed using commercial TaqMan assay reagents (Invitrogen Life Technologies).

Depletion of CXCR3L from serum

Aliquots (1 ml) of patient serum were incubated with 50 µl of protein G-Sepharose 4B (Sigma-Aldrich) for 1 h at room temperature under continuous agitation. For depletion of CXCR3L, 50 µl of Sepharose was washed in PBS and incubated with 5 µg of anti-CXCL9 (49106), anti-CXCL10 (33036), and anti-CXCL11 (87328) mAbs (all from R&D Systems) for 1 h at 4°C. Control mAb-loaded Sepharose was prepared in parallel. After several washes, mAb-bound Sepharose was added to precleared serum aliquots and incubated for 2 h under agitation. Sepharose was removed by centrifugation, and sera were stored at –80°C until use.

Transwell insert chemotaxis assay

Chemotaxis assay was performed, as described previously (16). Briefly, PBMC, either freshly isolated or cultured for 5 days, were washed and resuspended in migration buffer consisting of HBSS (Invitrogen Life Technologies), 1 mM CaCl2, 0.5 mM MgCl2, and 0.1% BSA (all from Sigma-Aldrich) at a density of 2 x 106 cells/ml. Chemokine solution (600 µl) was added to wells of a 24-well plate (Corning Glass). Buffer alone was used to determine spontaneous migration. Aliquots (100 µl) of resuspended cells were transferred to Costar Transwell inserts with 3-µm pore-size polycarbonate membranes (Corning Glass). Cells were allowed to migrate for 4 h at 37°C. Migrated cells were collected by complete aspiration of the content of the wells, followed by two washes of the wells. This rinsing fluid and the content of each well were pooled, and one-eighth (i.e., 100 µl) of each sample was used for cell counting on a FACScan. Absolute counts of migrated cells were obtained by multiplying the amount of cells found in the appropriate forward/sideward scatter gate by eight. Migration indices are calculated as follows: (percentage of chemokine-induced migration)/(percentage of spontaneous migration). In desensitization experiments, T cells were incubated with CXCR3- or mock-depleted serum for 1 h before the migration assay.

Quantification of serum chemokine levels

Serum was obtained from healthy donors and CTCL patients, and chemokines were measured using a cytometric bead assay (Human Chemokine Kit I; BD Biosciences), according to the instructions of the manufacturer. The assay was performed on a FACScan equipped with BD CBA software (BD Biosciences).

Quantification of phosphorylated ERK1/2

Freshly isolated CD8+ T cells of healthy donors and CTCL patients were exposed to CXCL11 (100 ng/ml; R&D Systems), anti-CD3 (OKT-3, 1 µg/ml), or medium only for 5 min. Cells were washed in chilled PBS and pelleted. Pellets were lysed in denaturation buffer (BD Biosciences) and boiled for 5 min. The total protein amount of each sample was determined using the bicinchoninic acid assay (Pierce Biotechnology). Equal amounts of total protein were subjected to quantification of ERK1/2 phosphorylation using a cytometric bead assay (PhosphoERK1/2 (T202/Y204) Flex Set; BD Biosciences), according to the instructions of the manufacturer. Data analysis was performed on a FACSCalibur equipped with the FACSComp software and analyzed with FCAP software (both from BD Biosciences).

Immunohistochemistry

Cryostat sections (5 µm) from CTCL tumors were fixed in acetone and stained with MECA-79 (IgM; BD Pharmingen) (17), anti-VCAM-1, anti-E-selectin Abs (18), or isotype-matched control mAbs. Bound Abs were visualized by biotinylated second-step Abs, peroxidase-labeled avidin biotin complexes, and amino-ethylcarbazol. Sections were counterstained with hematoxylin and coverslipped.

Statistical analysis

Data was analyzed by SPSS 12.0 (SPSS).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
PB T cells of CTCL patients down-regulate the surface expression of CXCR3, but not of other chemokine receptors

To identify possible migratory alterations of the major T cell subsets in PB of CTCL patients with hematologic involvement, we analyzed chemokine receptor expression in circulating normal CD7+CD4+ and CD8+ T cells as well as CD7–CD4+ lymphoma cells. As shown in Fig. 1, among 14 different CC and CXC chemokine receptors analyzed, selectively one chemokine receptor, CXCR3, is down-regulated in CD4+ and CD8+ T cells of CTCL patients as compared with the respective T cell subsets of healthy controls. In healthy subjects, CXCR3 is the chemokine receptor most abundantly expressed in CD8+ T cells (>70% of cells; Fig. 1) and is functionally associated with the ability of CD8+ T cells to extravasate and to exert cytotoxic functions in tissues (19, 20). In CTCL patients’ blood, only a minor subset of CD8+ T cells expresses CXCR3 (Fig. 1) and the level of CXCR3 expression falls below that of cells from control subjects (data not shown). In contrast, other chemokine receptors that are expressed by CD8+ T cells (e.g., CCR5, CCR7, CXCR1) (20, 21, 22, 23) showed no significantly altered expression in the T cells of CTCL patients. Circulating CD7–CD4+ CTCL lymphoma cells displayed only low levels of CXCR3, but expressed abundantly the chemokine receptors for homing to lymph nodes (CCR7) and skin (CCR4; Fig. 1). We noted a trend for increased CCR4 expression in CD4+ and CD8+ T cells of patients as compared with controls, but this difference did not reach statistical significance (Fig. 1, p > 0.05).


Figure 1
View larger version (20K):
[in this window]
[in a new window]

  		FIGURE 1. Comparative analysis of chemokine receptor expression on normal CD4+ T cells and CD8+ T cells from CTCL patients (n = 4) and healthy donors (n = 3). Values represent percentages of CD4+ T cells (left panel) and CD8+ T cells (center panel) that are reactive with mAbs specific for the indicated chemokine receptors. Right panel, Shows the chemokine receptor profile of circulating lymphoma cells from the same CTCL patients. Data were obtained by gating for CD3+CD4+CD7+ (left panel), CD3+CD8+ (center panel), and CD3+CD4+CD7– PBMC (right panel) using four-color flow cytometry. Asterisks indicate significant statistic differences in chemokine receptor expression between corresponding cell populations as calculated using Student’s t test (**, p ≤ 0.001).

 
Down-regulation of CXCR3 expression is associated with the degree of tumor burden, but is not directly related to therapy

We next asked whether the phenomenon of CXCR3 down-regulation in T cells of CTCL patients is associated with specific treatment regimens or, alternatively, is linked to the pathophysiology of the disease process in CTCL. Current immunomodulatory therapies for medium to far advanced CTCL include ECP, systemic administration of IFN-{alpha}, or a combination of both. CTCL patients were grouped according to the therapeutic regimen used and analyzed for CD8+ T cell CXCR3 expression. Treatment controls included patients who underwent ECP as an experimental therapy of systemic scleroderma and patients who received IFN-{alpha} in an adjuvant setting after surgery for primary cutaneous melanoma. Although we noted a trend for lower CXCR3 expression in CD8+ T cells of these treatment controls as compared with those of healthy nontreated subjects, this difference did not reach statistical significance (p > 0.05; Fig. 2A). In contrast, CXCR3 expression in CD8+ T cells was significantly reduced in CTCL patients as compared with nontreated controls (p < 0.05) and in most cases was below that of the control treatment groups (Fig. 2A). Thus, the extent and the invariable occurrence of CXCR3 down-regulation in CD8+ T cells of CTCL patients, but not in control treatment groups, strongly suggest a direct relation of CXCR3 down-regulation to the disease process itself, although a certain contribution of treatment cannot be ruled out.


Figure 2
View larger version (13K):
[in this window]
[in a new window]

  		FIGURE 2. Low surface expression of CXCR3 on CD8+ T cells in CTCL patients is correlated with disease extent and is not a direct consequence of therapy. A, Lack of correlation between the expression of CXCR3 on CD8+ T cells and therapy. The percentage of CXCR3-expressing CD8+ T cells (y-axis) from controls (healthy donors (no Tx, n = 5), patients suffering from systemic scleroderma undergoing ECP (n = 5), patients treated with IFN-{alpha} after surgery for primary melanoma (IFN-{alpha}) (n = 4), and from CTCL patients treated with ECP (n = 7), with IFN-{alpha} (n = 1), or with ECP and IFN-{alpha} (n = 3) was assessed by flow cytometry. Asterisk indicates a significant difference in CXCR3 expression between groups (p ≤ 0.05). B, Inverse correlation between the percentage of CD7–CD4+ circulating CTCL cells (x-axis) and the percentage of CXCR3-expressing CD8+ T cells (y-axis, p < 0.05). The curve depicts the logarithmic regression between the two data sets from patients ({circ}, n = 15) and healthy donors (•, n = 14). The dotted line demarcates the upper limit of the range in which CD7–CD4+ T cells are found in healthy donors. In the patients with CD7–CD4+ T cell counts above this level, T cell clonality was detected in the PB (data not shown).

 
To analyze this possibility further, we asked whether the level of CXCR3 down-regulation correlates with the extent of disease. As a measure of tumor burden, we determined the numbers of tumor cells in the PB of CTCL patients. This parameter is justified by the fact that progression from local to systemic CTCL is almost invariably associated with the appearance of circulating tumor cells and the extent of systemic disease is closely reflected by the tumor cell counts in blood (2). Accordingly, we included CTCL patients with low to high tumor cell counts in the blood (>10% CD7–CD4+ circulating T cells) as well as patients who suffered from skin-restricted disease (<10% CD7–CD4+ T cells). This analysis revealed a statistically significant inverse correlation between the level of CXCR3 expression on CD8+ T cells and the percentage of CD7–CD4+ tumor cells in PB (R2 > 0.5; Fig. 2B). No statistically significant correlation between CXCR3 expression and total or relative counts of CD7+CD4+ and CD8+ T cells was observed (R2 < 0.5; data not shown). Other chemokine receptors of the CD8+ T cells (i.e., CCR5, CCR7, CXCR1) failed to show a similar disease extent-related regulation (data not shown). Thus, it appears that the degree of CXCR3 down-regulation in CD8+ T cells is linked to the magnitude of the lymphoma cell burden, but not to a lesser degree to CTCL-related alterations in the homeostasis of normal T cells. Accordingly, we focused our experimental attention to the resolution of the mechanism of CXCR3 down-regulation in CTCL patients with substantial hematologic involvement.

Down-regulation of CXCR3 on CD8+ T cells in CTCL is not linked to inefficient generation and function of CTL

On the basis of their surface receptor expression, CD3+CD8+ T cells can be subdivided into naive T cells (CD27+CD45RA+), memory T cells (CD27+CD45RA–), and effector T cells (CD27–CD45RA+); a fourth subset of CD27–CD45RA– cells can be resolved. It corresponds to CD8+ cells that are in transition between the memory and the effector cell pool (Fig. 3A) (21). Because CXCR3 is of particular relevance for the migration of cytotoxic effector CD8+ T cells, we asked whether the down-regulation of CXCR3 expression in CD8+ T cells of CTCL patients can be explained by a reduction of effector T cells in PB. Intriguingly, our comparative analysis of the CD8+ T cell subset composition in CTCL patients and healthy controls revealed that this is not the case. Conversely, effector CD8+ T cells were enriched among CD8+ T cells in CTCL patients (Fig. 3, A and B). In absolute numbers, effector CD8+ T cells of both the CD27–CD45RA+ and the CD27–CD45RA– phenotype occurred in similar prevalence in patients and controls (Fig. 3C). Thus, the previously observed striking reduction in PB-CD8+ T cell counts in advanced CTCL (10) is due to a depletion of naive as well as of memory CD8+ T cells (Fig. 3C).


Figure 3
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 3. Depletion of naive and memory CD8+ T cells, but persistence of effector CD8+ T cells in PB of CTCL patients. A, Analysis of CD8+ T cell differentiation in CTCL patients (right panel) and healthy donors (left panel). Functional CD8+ T cell subpopulations (Tn, naive CD8+ T cells; Tm, memory CD8+ T cells; Te, effector CD8+ T cells) were identified by their differential CD27 (y-axis) and CD8+ CD45RA expression (x-axis). CD8+ T cells were gated as CD3+CD8+ PBMCs in four-color immunocytometry analyses. Quadrants in dot plots are set according to the reactivity of isotype-matched control mAbs. Numbers in the inset indicate quadrant statistics. B, Mean percentages of CD8+ Tn, Tm, and Te subsets among all CD8+ T cells in PB of healthy donors (n = 5) and CTCL patients (n = 2). Tm/e cells are the CD27– and CD45RA-negative CD8+ T cells, which are in transition between the memory and the effector cell pool. Analyzed patients had a tumor load of 18.5–21.4% CTCL cells in blood. C, Absolute cell counts of CD8+ T cell subsets in the blood of healthy donors and CTCL patients. Values were obtained by multiplying the percentages of the individual subpopulations with the absolute number of CD8+ T cells/ml blood. Error bars in B and C indicate SD.

 
In support of an unperturbed CD8+ effector T cell function in CTCL patients’ blood is the finding of enhanced rather than reduced activation marker expression in these cells (Fig. 4A). Furthermore, the cytotoxic potential of patient-derived CD8+ T cells was found to be at least comparable to that of healthy donors’ CD8+ T cells (Fig. 4B). When we sorted patient-derived CD8+ T cells according to their naive, memory, and effector phenotypes and used them in a sensitive Europium-based cytotoxicity assay, we observed that cytotoxic function was clearly enriched in the immunophenotypically defined effector cell populations (Fig. 4C), which is comparable to the situation seen in nondiseased individuals (data not shown) (21). When we comparatively analyzed CXCR3 expression in naive, memory, and effector CD8+ T cells of CTCL patients and controls, we observed that CXCR3 down-regulation is most pronounced in the effector CD8+ T cell population of patients (Fig. 4D). Thus, we conclude that CXCR3 down-regulation in PB-CD8+ T cells of CTCL patients mainly involves cytotoxic effector T cells. These cells, however, neither are compromised in their cytotoxic potential nor are reduced in numbers.


Figure 4
View larger version (23K):
[in this window]
[in a new window]

  		FIGURE 4. CD8+ T cells from CTCL patients display an activated phenotype and are not compromised in their cytotoxic potential. A, Expression of activation markers on CD8+ cells. Given are percentages of total CD8+CD3+ T cells expressing the indicated activation Ags in CTCL patients (n = 9; {square}) and healthy donors (n = 5; {blacksquare}) as determined by flow cytometry. The asterisk indicates significant statistic difference as determined by Student's t test (*, p ≤ 0.05). B, Cytotoxic potential of CD8+ T cells of CTCL patients and healthy donors. K-562 cells were used as targets for purified CD8+ T cells (purity >95%) from CTCL patients ({square}, n = 2) and healthy controls ({blacksquare}, n = 2). Anti-CD3 mAbs were bound to Fc{gamma}Rs on K-562 to mediate TCR/CD3-induced T cell activation upon contact with target cells. Percentages of target cells killed by CD8+ T cells were measured by flow cytometry. Mean values of experiments performed in triplicates are given. Error bars indicate SD. C, Cumulative cytotoxicity of functional CD8+ T cell subsets in CTCL patients in a 4-h Eu3+-release assay. CD8+ T cell subsets (Tn, naive CD8+ T cells; Tm, memory CD8+ T cells; Tm/e, memory/effector CD8+ T cells; Te, effector CD8+ T cells) were flow sorted according to their differential CD27 and CD45RA expression and assayed in an anti-CD3-based, redirected Eu3+-release assay for cytotoxicity against K-562 cells. Specific lysis was calculated as the difference of lysis of nonlabeled targets and anti-CD3-labeled targets. Mean values (+SD) of triplicate determinations are given. Representative experiment. D, Down-regulation of CXCR3 expression on CD8+ T cells in CTCL patients is most prominent in the cytotoxic effector T cell population. The expression of CXCR3 was measured on CD27/CD45RA-defined CD8+ T cell subsets of patients (n = 2) and healthy donors (n = 5) by flow cytometry. Relative CXCR3 expression on CD8+ T cell subsets of CTCL patients vs healthy donors was calculated by forming the quotient of the mean percentages from both groups of donors and for the indicated T cell subsets. Error bars represent SD.

 
Down-regulated surface expression of CXCR3 on CD8+ T cells is associated with the accumulation of the receptor in endolysosomal compartments

To see whether the selective defect in CXCR3 expression in T cells in CTCL patients is due to transcriptional silencing, we performed quantitative PCR experiments on purified CD8+ T cells. As shown in Fig. 5A, normalization of the amount of CXCR3 transcripts for housekeeping gene (beta2m) expression revealed that CXCR3 transcripts in freshly isolated CD8+ T cells of CTCL patients are not reduced, but rather augmented, as compared with healthy donors. Culture of the isolated CD8+ T cells in the presence of IL-2 slightly up-regulated CXCR3A in cells from controls, but not patients (Fig. 5A). The migration-competent A form of CXCR3 (CXCR3A) was by far the most prevalent CXCR3 transcript expressed in CD8+ T cells of patients and controls (Fig. 5A). mRNAs for the splice variant CXCR3B, a receptor that binds platelet factor-4 as well as classical CXCR3L without migration-inducing activity (15), and a truncated version of CXCR3 comprising only four transmembrane domains (22) were expressed at comparatively low abundance (Fig. 5A, data not shown). To analyze whether CD8+ T cells of CTCL patients display impaired translation of CXCR3, we comparatively analyzed surface expression and total cellular expression of CXCR3. Although down-regulated at the cell surface, CXCR3 protein was present intracytoplasmically in CD8+ T cells of CTCL patients (Fig. 5B). Thus, relocation of CXCR3 to intracellular sites, but not the impairment in transcription and translation is responsible for down-regulated cell surface expression in CD8+ T cells of CTCL patients.


Figure 5
View larger version (43K):
[in this window]
[in a new window]

  		FIGURE 5. Down-regulated T cell CXCR3 expression in CTCL results from relocation of CXCR3 from the surface to endosomal organelles. A, CXCR3A and CXCR3B mRNA expression in purified CD8+ T cells of CTCL patients (n = 3) and healthy donors (HD, n = 3). The relative amount of CXCR3 mRNA was measured by real-time PCR and normalized to beta2m expression. f, Freshly isolated cells; c, cells cultured for 5 days in medium containing IL-2. Mean values (+SD) of three experiments are given. B, Reduced surface, but high total cellular CXCR3 protein expression in CD8+ T cells of CTCL patients. CXCR3 expression was measured by flow cytometry on PBMC labeled with anti-CD3 allophycocyanin and anti-CD8 PE before (left panels) or after fixation and permeabilization (right panels). mAb reactivity with CXCR3 is shown by filled histograms, and isotype control reactivity by open histograms. C, Intracellular localization of CXCR3 in CD8+ T cells from CTCL patients. Purified CD8+ T cells were mounted on slides, fixed, permeabilized, and immunostained with anti-CXCR3 (Alexa488, green) and anti-CD3 (Cy5, red). Merge of the emissions (yellow) demonstrates colocalization of CXCR3 and CD3 on the surface of CD8+ T cells from healthy donors (bottom panel). In CD8+ T cells from CTCL patients, CD3 surface expression is maintained, but CXCR3 accumulates in vesicular structures beneath the surface membrane (arrows, top panel). Images shown are representative for results obtained with CD8+ T cells from three different CTCL patients. D, Intracellular CXCR3 is contained in EEA1+ and EEA1+LAMP2+ vesicular compartments. Purified CD8+ T cells from CTCL patients were mounted on slides, fixed, permeabilized, and immunostained with anti-CXCR3 PE (red), anti-LAMP2 FITC (green), and anti-EEA1 (allophycocyanin, blue). EEA1+LAMP2– endosomes containing CXCR3 appear in magenta (white arrows), whereas EEA1+LAMP2+ endolysosomal compartments are white (gray arrows) in the overlay exposure. E, Serum levels of CXCR3L. Sera were obtained from CTCL patients (n = 5) and healthy donors (n = 5). The amounts of the indicated chemokines in the serum were determined using cytometric bead assays. Error bars represent SD. The asterisk indicates significant statistic difference as determined using Student’s t test (*, p ≤ 0.05). F, Rescue of CXCR3 expression on CD8+ T cells from CTCL patients by in vitro culture. CXCR3 surface expression was measured on CD8+ T cells, freshly isolated (left panels), or cultured for 5 days in medium containing IL-2. Dot plots (CD3, y-axis; CXCR3, x-axis) show cells gated for reactivity with anti-CD8 mAbs. Quadrants are set according to isotype control mAb reactivities.

 
To better define the cytoplasmic localization of CXCR3 in CD8+ T cells in CTCL patients, we performed image analyses using confocal microscopy. In CD8+ T cells from healthy donors, CXCR3 was expressed almost exclusively at the cell surface, as evidenced by colocalization with cell surface-expressed TCR/CD3 (Fig. 5C). In striking contrast, CXCR3 in CTCL patients localized in vesicular structures juxtaposed to but clearly demarcated from cell membrane (Fig. 5C). Triple stainings for the endosomal marker EEA1 and the lysosomal Ag LAMP2 revealed that CXCR3 protein is present in endosomes (EEA+LAMP2–), lysosomes (EEA–LAMP2+), and EEA+LAMP2+ compartments that may represent endolysosomes (Fig. 5D). Thus, the reduced CXCR3 surface expression in T cells of CTCL patients can be the result of endosomal retention or lysosomal degradation or both. To investigate whether soluble CXCR3L may mediate the reduced CXCR3 surface expression in CTCL patients, we measured serum levels of the CXCR3L CXCL10 and CXCL9 in patients and healthy controls. A statistically significant increase in CXCL10 serum levels was observed in CTCL patients’ sera, whereas the increase in CXCL9 levels did not reach statistical significance (Fig. 5E). To see whether CXCR3 down-modulation in vivo can be reverted in vitro, we cultured CD8+ T cells under serum-free medium conditions. Although control CD8+ T cells only moderately up-regulated CXCR3, CD8+ T cells from CTCL patients strongly increased CXCR3 surface expression and reached expression levels close to those seen in controls (Fig. 5F).

CXCR3L fail to induce chemotaxis, but activate ERK in CD8+ T cells from CTCL patients

CXCR3 can activate at least two distinct signaling pathways in T cells, as follows: Gi proteins necessary for migration and the ERK/MAPK pathway involved in the regulation of survival. To see whether down-regulated CXCR3 surface expression in CD8+ T cells from CTCL patients results in functionally impaired responses to CXCR3L, purified CD8+ T cells from CTCL patients and healthy controls were subjected to Transwell migration and ERK activation assays. As shown in Fig. 6A, freshly isolated control CD8+ T cells exhibited weak, but reproducible chemotactic responses to CXCR3L. CXCL11 was used in these experiments because it induces stronger T cell migration than the other CXCR3L in vitro (23). In contrast to CD8+ T cells from controls, CD8+ T cells from CTCL patients were deficient in their chemotactic response to CXCR3L (Fig. 6A). In vitro culture of CD8+ T cells from CTCL patients resulted in recovery of chemotactic activity to CXCR3L (Fig. 6A), which is in accordance with the observed rescue of CXCR3 surface expression. The magnitude of response was, however, invariably lower with cultured CD8+ T cells from CTCL patients than from healthy controls. In contrast, CD8+ T cells from patients and from controls were equally responsive to CCR5 ligands (Fig. 6B) and nonresponsive to the CXCR3B ligand PF-4 (data not shown). Although incapable of migrating to CXCR3L, fresh CD8+ T cells from CTCL patients showed vigorous phosphorylation of ERK upon exposure to CXCR3L (Fig. 6C). The level of phosphorylation of ERK induction by CXCR3 triggering was even comparable to that seen after optimized TCR/CD3 triggering (Fig. 6C). Thus, the redistribution of CXCR3 to endolysosomal compartments of CD8+ T cells in CTCL patients abolishes complex ligand-induced cellular processes, such as migration, whereas ERK activation remains functional.


Figure 6
View larger version (37K):
[in this window]
[in a new window]

  		FIGURE 6. Partial loss of CXCR3 functionality in T cells from CTCL patients. A, Circulating T cells from CTCL patients fail to migrate to CXCR3L. Freshly isolated and cultured CD8+ (left panel) and CD4+ T cells (right panel) from CTCL patients ({diamond}) and healthy donors ({blacksquare}) were subjected to Transwell chemotaxis assays. Cells were assayed for chemotaxis to graded concentrations of CXCL11 (x-axis). Shown are mean migration indices (±SD; y-axis) as obtained in experiments with two donors per group in triplicates. Spontaneous migration was similar with T cells from patients and controls. B, Circulating T cells from CTCL patients and healthy donors migrate to CCR5 ligands. Freshly isolated CD8+ (left panel) and CD4+ T cells (right panel) from CTCL patients ({diamond}) and healthy donors ({blacksquare}) were assayed for chemotactic responsiveness to the CCR5 ligand MIP-1{alpha}. Shown are mean migration indices (±SD; y-axis) obtained in experiments with two donors per group in triplicates. C, CXCR3L induce ERK1/2 phosphorylation in T cells from CTCL patients: CD8+ T cells of CTCL patients (purity >95%) were incubated with medium only, anti-CD3 mAbs (OKT-3, 1 µg/ml; concentration optimized for peak ERK1/2 phosphorylation), or CXCL11 (100 ng/ml). The cellular content of phosphorylated ERK1/2 (at positions T202 and Y204) was measured using a cytometric bead assay, and data were normalized to total cellular protein. Shown are mean increases (±SEM) in normalized phosphorylated ERK1/2 over medium control (Med) as obtained in experiments with three donors in duplicates. D, CXCR3L in serum of CTCL patients can reduce CXCR3-mediated chemotaxis of healthy donors’ T cells. IL-2-activated T cells of healthy donors were incubated with CXCL11-, CXCL9-, and CXCL10 (CXCR3L)-depleted or mock-depleted serum (40%, v/v) from CTCL patients. Then cells were tested in Transwell chemotaxis assays for migratory responsiveness to graded concentrations of CXCL11 (x-axis). Diluted serum remained in the upper chamber of the Transwell device during the assay. Spontaneous T cell migration was not influenced by preincubation of T cells with either CXCR3L- or mock-depleted serum (data not shown). Shown are mean migration indices (+SD; y-axis) of experiments performed in triplicates. Asterisk indicates significant statistic difference as determined by Student’s t test (*, p ≤ 0.05).

 
CXCR3 is down-regulated not only in CD8+, but also in CD4+ T cells of CTCL patients (Fig. 1). Thus, we analyzed whether CD4+ T cells in CTCL have an impaired ability to migrate to CXCR3L. As shown in Fig. 6A, freshly isolated CD4+ T cells of CTCL patients do not migrate to CXCR3L, arguing for a down-regulation of CXCR3 function in the major circulating T cell subsets. The further observation that serum levels of CXCR3L are significantly increased in CTCL patients prompted us to investigate whether ligand-induced receptor down-regulation in vivo may contribute to the observed migratory defect. To address this possibility directly, sera from CTCL patients were depleted or were not depleted of CXCR3L by immunoaffinity chromatography. Normal T cells were incubated in CXCR3L- or mock-depleted patient serum before allowing them to migrate to CXCL11. As shown in Fig. 6D, CXCR3L in CTCL patients’ sera are capable of reducing CXCR3 functionality because T cell chemotaxis to CXCL11 was significantly increased upon removal of CXCR3L. Thus, contact with CXCR3L in serum is at least partially responsible for down-regulation of CXCR3 in T cells from CTCL patients.

Impairment of CXCR3 function prevents the recruitment of circulating effector CD8+ T cells into tumor lesions

The consequence of a loss of CXCR3-dependent T cell migration in CTCL patients could be the inability of circulating CD8+ T cells to extravasate into neoplastic tissue and to exert cytotoxic function. To directly address this possibility, we comparatively analyzed the subset composition of CD8+ T cells in the blood and in tumor lesions of three CTCL patients. In three of three patients, we observed only few effector CD8+ T cells in the lesions, whereas the same patients had a substantial population of effector CD8+ T cells in the blood (Fig. 7). Poor detectability of effector T cells in tissues is unlikely the reason for these observations because effector T cells of the memory/effector CD8+ T cell phenotype were the most prevalently enriched T cell population detected in virus-induced skin lesions (Fig. 7B). In contrast to effector T cells, memory CD8+ T cells that express CCR5 as their cardinal chemokine receptor (24, 25) (data not shown) were efficiently recruited into CTCL lesions (Fig. 7B). Interestingly, the memory CD8+ T cell population in CTCL lesions compared with that of blood contained more CCR5-, but not CCR7- or CXCR3-expressing cells (CCR5+, +32.7%; CCR7+, +1.4%; CXCR3+, –10.9%; means of two independent experiments), as revealed by five-color flow cytometry. This further suggests that CCR5 is a critical receptor for memory T cell recruitment into CTCL lesions. In contrast, the very few effector CD8+ T cells found in CTCL lesions showed higher CXCR3, but not CCR5 or CCR7 expression than effector CD8+ T cells in blood (CXCR3+, +12%; CCR5+, +1.4%; CCR7+, –10.9%). Thus, it appears that the impairment of CXCR3-dependent functionality in CTCL results in dramatically lowered effector CD8+ T cell recruitment into tumors, and that the few effector CD8+ T cells in CTCL tumors are recruited on the basis of residual CXCR3 functionality.


Figure 7
View larger version (31K):
[in this window]
[in a new window]

  		FIGURE 7. Absence of CD8+ effector T cells, but presence of naive and memory CD8+ T cells in CTCL skin tumors. A, T cells from blood (left panels) and skin tumor lesions (right panels) of CTCL patients (n = 3) were isolated and subjected to four-color flow cytometry. CD3+CD8+ were gated and functional CD8+ T cell subsets were resolved by their differential expression of CD27 (y-axis) and CD45RA (x-axis, upper panels). Data shown in the lower panels represent mean percentages (+SD) of subset representation among all CD8+ T cells from blood (left panel) and skin tumors (right panel) as obtained in experiments with cells from three CTCL patients. B, Differential recruitment of CD8+ T cell subpopulations to CTCL- and virus-induced skin lesions. The data from the lower panels of A were expressed as quotient of subset frequencies from tumor and blood (left panel). In this representation, a value >1 denotes relative enrichment of a given CD8+ T cell subset in the skin lesions, whereas a value <1 indicates a relative deficit in recruitment to the skin lesion. The same calculation was performed for CD8+ T cells isolated from blood and skin lesions of a patient infected with VZV (right panel).

 
Defective effector CD8+ T cell homing to CTCL lesions is not due to the absence of CXCR3L within the lesions

To see whether the specific ligands of CXCR3 are present within CTCL lesions, we performed immunofluorescence analyses on sections from excised skin tumor material. As shown in Fig. 8A, anti-CXCL9 and anti-CXCL11 immunoreactivity was associated with the majority of the blastoid cells forming CTCL skin tumors. To exclude that infiltrating cells other than T cells (i.e., dendritic cells, macrophages, and B cells) may significantly contribute to the observed CXCL9 and CXCL11 expression in CTCL lesions, we performed double stainings using anti-MHC class II mAbs (Fig. 8A). Results obtained strongly suggest that CXCR3L expression is largely restricted to MHC class II-negative CTCL cells. Interestingly, CXCL10 was barely detected in tumor cells, but was present in granular cells within the lesions, most likely representing mast cells (data not shown). Thus, the defective effector CD8+ T cell recruitment to CTCL lesions cannot be attributed to a lack of CXCR3L expression in CTCL lesions. Conversely, CXCR3L production in CTCL tumor lesions, most likely by tumor cells themselves, may contribute to the observed increased serum sevels of these cytokines in patients with advanced CTCL.


Figure 8
View larger version (123K):
[in this window]
[in a new window]

  		FIGURE 8. Expression of chemokines and vascular adhesion molecules in CTCL skin tumors. A, CXCR3L are expressed in CTCL skin tumors. Confocal microscopy immunofluorescence analysis reveals CXCL9 (left panel) and CXCL11 protein expression (right panel) in CTCL lesions. CXCR3L expression was found associated with blastoid tumor cells (green, upper panel), but not in MHC class II-expressing cells infiltrating the tumor (CXCR3L, red; MHC class II, green). No staining was detected with isotype control mAbs (inset, left upper panel) and with anti-CXCL9 and anti-CXCL11 mAbs on normal human skin (data not shown). Data shown are representative for results obtained with three CTCL tumor specimens from three different donors. B, In skin lesions of tumor stage CTCL, blood vessels of the dermis express peripheral node addressins as identified by MECA-79 reactivity, poorly express E-selectin, and are negative for VCAM-1. Serial sections of skin biopsies from CTCL skin tumors were stained with mAbs specific for peripheral node addressins/L-selectin ligands (mAb MECA-79), E-selectin, or VCAM-1, or isotype-matched control mAbs. The low magnification image shows pronounced vascular MECA-79 reactivity in a representative CTCL tumor lesion (left panel). The inset in A denotes the area on the section for which MECA-79 reactivity, VCAM-1, and E-selectin expression were analyzed on consecutive serial sections (right panels). The phenotype of CTCL tumor vessels clearly differs from that seen in inflammatory reactions of the skin, where blood vessels express E-selectin and VCAM-1, but not MECA-79-reactive moieties (data not shown).

 
Recruitment of memory and naive CD8+ T cells correlates with high L-selectin ligand and low E-selectin and VCAM expression on endothelial cells in CTCL tumors

To further investigate receptor-ligand pairs possibly involved in the observed memory and naive CD8+ T cell recruitment into CTCL skin tumors, CD8+ T cell subsets were analyzed for the expression of receptors for vascular addressins. Lesional memory and naive CD8+ T cells, but not effector CD8+ T cells, expressed substantial levels of L-selectin, whereas all cell populations expressed the E- and P-selectin ligand-1 (data not shown). The expression of L-selectin ligands and other vascular addressins in CTCL skin tumors was analyzed by immunohistochemistry. As shown in Fig. 8B, the tumor vasculature in CTCL expresses abundantly MECA-79-reactive L-selectin ligands. In contrast, the activation-regulated endothelial adhesion molecules E-selectin and VCAM-1 are expressed in trace amounts only. Control stainings on tonsillar lymphoid tissue revealed strong L-selectin ligand along with poor E-selectin and VCAM-1 expression on the vasculature of T cell zones (data not shown). Thus, the vascular addressin pattern in CTCL tumors is similar to that seen in normal lymphoid organs. High-level L-selectin ligand expression by endothelial cell in CTCL tumors hence appears functionally relevant and responsible for the observed recruitment of memory and naive CD8+ T cells into these lesions.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
CXCR3 is a prime receptor involved in targeting CD8+ cytotoxic effector T cells into the skin (19) and other organs (20). In this study, we show that CXCR3 is down-modulated from the surface of CD8+ cytotoxic effector T cells of CTCL patients with advanced disease. However, CXCR3 transcripts and total amounts of CXCR3 protein were not reduced in these cells. In patients’ CD8+ T cells, CXCR3 protein accumulates in endolysosomal organelles, a feature indicative of ligand-induced receptor internalization (26). Indeed, we found strong evidence for ligand-induced down-regulation of CXCR3 surface expression and function in advanced CTCL. In CTCL patients with systemic involvement, we observed the following: 1) increased serum levels of CXCR3L; 2) decreased expression of CXCR3, but not of other chemokine receptors on CD8+ and CD4+ T cells; 3) abolished migratory T cell responses to CXCR3L, but not to other chemokines; and 4) partial rescue of CXCR3-mediated T cell migration after ex vivo culture of patients’ T cells. Moreover, CXCR3L in patients’ serum were capable of strongly reducing CXCR3-dependent chemotaxis of normal T cells in Transwell migration assays.

Besides the defect in CXCR3 expression and function, we found rather unperturbed differentiation effector CD8+ T cells in CTCL patients. This is evidenced by the following findings: 1) purified CD8+ T cell from patients and control subjects had a similar cytotoxic potential; 2) CD8+ T cell with the immunophenotype of cytotoxic effector cells occurred in comparable absolute numbers in PB of patients and controls; and 3) these immunophenotypically defined effector T cells had a higher cytotoxic potential than memory and naive CD8+ T cells from the same patient. Moreover, unlike CXCR3, other chemokine receptors that are associated with cytotoxic effector cell differentiation, e.g., CXCR1 (27), were expressed at equal levels in patients and controls. Thus, we conclude that down-regulation of CXCR3 in CTCL patients is not associated with a global defect in CD8+ effector T cell generation and function. Our investigations on patients with hematologic involvement reveal that naive CD8+ T cells and memory CD8+ T cells are the subpopulations most explicitly depleted from blood of CTCL patients. Thus, drastically lowered CD8+ T cell counts in PB are not necessarily associated with the inability to generate and maintain a circulating functional effector CD8+ T cell pool in advanced CTCL. It is, however, important to mention that the observed relative increase in effector CD8+ T cells in patient blood is not necessarily due to an augmented efficacy in effector CD8+ T cell generation. Our data suggest that effector CD8+ T cells, but not the other CD8+ T cell subsets, have an impaired ability to extravasate. Thus, effector CD8+ T cells are trapped in the circulation, which may further explain the well known propensity of CTCL patients to acquire infectious diseases with a severe clinical course (28).

To address directly the biological impact of CXCR3 down-regulation on effector CD8+ T cell homing to tumor sites, we comparatively analyzed blood and CTCL skin tumors. Tumorous skin lesions were largely devoid of effector CD8+ T cells. Because migration to CXCR3L was the only chemokine-dependent migratory activity deregulated in patients’ effector CD8+ T cells, one could speculate that effector CD8+ T cell homing is a nonredundant CXCR3-dependent mechanism (19, 20, 29). Interestingly, naive CD8+ T cells along with memory CD8+ T cells migrated in CTCL tumors. The endothelium of CTCL lesions fails to express activation-associated E-selectin and VCAM-1, but expresses ligands for L-selectin (17), a feature shared with high endothelial venules of normal lymphoid tissue. Our observation of naive T cell recruitment to CTCL tumors strongly suggests that L-selectin-binding moieties on CTCL tumor vessels are functionally relevant in vivo. The lack of E-selectin on these vessels may further contribute to the observed defect in effector CD8+ T cell homing. This situation is contrasted by the preferential homing of CXCR3+ effector CD8+ T cells to virus-induced skin lesions with CXCR3L presentation on E-selectin+ blood vessels (Fig. 7B) (30). Thus, modulation of endothelial selectin expression and down-regulation of effector CD8+ T cell migration are two functionally distinct, but synergistically acting mechanisms of CTCL escape from cytotoxic T cell attack.

As the result of our study, one can argue that the prevention or the reversal of the migratory defect of effector CD8+ T cells is a valid therapeutic option in advanced CTCL. IFN-{alpha}, currently used in CTCL, is a major inducer of CXCR3-binding chemokines. Thus, IFN-{alpha} may induce two effects with opposing outcome. On the one hand, IFN-{alpha}, when expressed locally, can promote CXCR3-mediated effector CD8+ T cell recruitment. In contrast, IFN-{alpha}, when inducing high serum levels of CXCR3L, may also contribute to down-regulation of effector CD8+ T migration. However, IFN-{alpha} treatment of control patients never resulted in CXCR3 down-regulation to the extent seen in IFN-{alpha}-treated CTCL patients. Thus, IFN-{alpha} treatment alone cannot explain the CXCR3 defect observed in CTCL patients. Increased serum levels of CXCR3L in CTCL patients could be also due to secretion from lymphoma cells autonomously, which is supported by the immunohistochemical detection of CXCR3L in CTCL tumors. Despite the fact that it is a plausible explanation, we currently cannot unequivocally prove that the increased serum levels of CXCR3L derive from the tumor cells exclusively. It is conceivable that mechanisms other than chemokine production contribute to high systemic levels of CXCR3L in CTCL patients, an argument supported by the observations of increased CXCL10 serum levels, but no apparent overexpression of this chemokine in the tumors. Accordingly, it is possible that serum levels of CXCR3L are further regulated by CTCL-associated alterations in chemokine metabolism. The enzyme dipeptidyl peptidase IV/CD26, which can initiate the digestion of CXCR3L, is expressed by all normal T cells, but is notoriously absent from CTCL cells (31). Reduction of normal T cells along with an increase in CTCL cells in blood may thus result in decreased catabolism of CXCR3L and effective receptor down-regulation by excess soluble chemokine as the consequence. This may also explain the correlation between CTCL cell counts in blood and the level of CXCR3 down-modulation on normal T cells in our patients. Thus, selective enzymatic or affinity-based removal of desensitizing chemokines from blood could be an interesting strategy to promote effector T cell migration in the setting of cancers that escape immune surveillance by exploitation of chemokine-dependent immune cell migration.

In summary, our findings highlight a new mechanism by which CTCL can manipulate the immune system to evade T cell-mediated destruction. Interestingly, in advanced metastatic melanoma, down-regulation of CXCR3 on CD8+ T cells was identified as a predictor of a poor clinical outcome (32). Thus, it appears that histogenetically distinct cancers manipulate CXCR3 expression and function in T cells to lock themselves from effector T cell attack. The development of specific interference strategies against these cancer-induced alterations of effector T cell homing can be an important requirement for the design of cancer vaccines that are clinically effective.


   Acknowledgment
 
We thank B. Reininger for excellent technical assistance.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
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.

1 This work was supported by the Research Center for Molecular Medicine (CeMM) of the Austrian Academy of Sciences, the Austrian Science Foundation (Fonds zur Forderung der Wissenschaftlichen Forschung; Projects SFB F1813 and SFB F2308), and the Genome Research in Austria (GEN-AU) program of the Austrian Ministry of Science (to D.M.). Back

2 Address correspondence and reprint requests to Dr. Dieter Maurer, Department of Dermatology, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria. E-mail address: dieter.maurer{at}meduniwien.ac.at Back

3 Abbreviations used in this paper: CTCL, cutaneous T cell lymphoma; 7-AAD, 7-aminoactinomycin D; beta2m, beta2-microglobulin; CXCR3L, CXCR3 ligand; EEA, early endosomal Ag; ECP, extracorporeal photochemotherapy; LAMP, lysosome-associated membrane protein; PB, peripheral blood; VZV, varicella zoster virus. Back

Received for publication September 8, 2006. Accepted for publication July 2, 2007.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Harmon, C. B., T. E. Witzig, J. A. Katzmann, M. R. Pittelkow. 1996. Detection of circulating T cells with CD4+CD7– immunophenotype in patients with benign and malignant lymphoproliferative dermatoses. J. Am. Acad. Dermatol. 35: 404-410. [Medline]
  2. Laetsch, B., A. C. Häffner, U. Dbbeling, B. Seifert, E. Ludwig, G. Burg, R. Dummer. 2000. CD4+/CD7– T cell frequency and polymerase chain reaction-based clonality assay correlate with stage in cutaneous T cell lymphomas. J. Invest. Dermatol. 114: 107-111. [Medline]
  3. Bernengo, M. G., M. Novelli, P. Quaglino, F. Lisa, A. De Matteis, P. Savoia, N. Cappello, M. T. Fierro. 2001. The relevance of the CD4+ CD26– subset in the identification of circulating Sézary cells. Br. J. Dermatol. 144: 125-135. [Medline]
  4. Matutes, E., D. Robinson, M. O’Brien, B. F. Haynes, H. Zola, D. Catovsky. 1983. Candidate counterparts of Sézary cells and adult T-cell lymphoma-leukemia cells in normal peripheral blood: an ultrastructural study with the immunogold method and monoclonal antibodies. Leuk. Res. 7: 787-801. [Medline]
  5. Siegel, R. S., T. Pandolfino, J. Guitart, S. Rosen, T. M. Kuzel. 2000. Primary cutaneous T-cell lymphoma: review and current concepts. J. Clin. Oncol. 18: 2908-2925. [Abstract/Free Full Text]
  6. Winter, D., E. Fiebiger, P. Meraner, H. Auer, C. Brna, R. Strohal, F. Trautinger, R. Knobler, G. F. Fischer, G. Stingl, D. Maurer. 2003. Definition of TCR epitopes for CTL-mediated attack of cutaneous T cell lymphoma. J. Immunol. 171: 2714-2724. [Abstract/Free Full Text]
  7. Eichmüller, S., D. Usener, R. Dummer, A. Stein, D. Thiel, D. Schadendorf. 2001. Serological detection of cutaneous T-cell lymphoma-associated antigens. Proc. Natl. Acad. Sci. USA 98: 629-634. [Abstract/Free Full Text]
  8. Eichmüller, S., D. Usener, D. Thiel, D. Schadendorf. 2003. Tumor-specific antigens in cutaneous T-cell lymphoma: expression and sero-reactivity. Int. J. Cancer 104: 482-487. [Medline]
  9. Häffner, A. C., A. Tassis, K. Zepter, M. Storz, O. Tureci, G. Burg, F. O. Nestle. 2002. Expression of cancer/testis antigens in cutaneous T cell lymphomas. Int. J. Cancer 97: 668-670. [Medline]
  10. Vermeer, M. H., R. van Doorn, D. Dukers, M. W. Bekkenk, C. J. Meijer, R. Willemze. 2001. CD8+ T cells in cutaneous T-cell lymphoma: expression of cytotoxic proteins, Fas ligand, and killing inhibitory receptors and their relationship with clinical behavior. J. Clin. Oncol. 19: 4322-4329. [Abstract/Free Full Text]
  11. Hoppe, R. T., L. J. Medeiros, R. A. Warnke, G. S. Wood. 1995. CD8-positive tumor-infiltrating lymphocytes influence the long-term survival of patients with mycosis fungoides. J. Am. Acad. Dermatol. 32: 448-453. [Medline]
  12. Willemze, R., E. S. Jaffe, G. Burg, L. Cerroni, E. Berti, S. H. Swerdlow, E. Ralfkiaer, S. Chimenti, J. L. Diaz-Perez, L. M. Duncan, et al 2005. WHO-EORTC classification for cutaneous lymphomas. Blood 105: 3768-3785. [Abstract/Free Full Text]
  13. Kriehuber, E., W. Bauer, A. S. Charbonnier, D. Winter, S. Amatschek, D. Tamandl, N. Schweifer, G. Stingl, D. Maurer. 2005. The balance between NF-{kappa}B and JNK/AP-1 activity controls dendritic cell life and death. Blood 106: 175-183. [Abstract/Free Full Text]
  14. Blomberg, K., C. Granberg, I. Hemmila, T. Lovgren. 1986. Europium-labelled target cells in an assay of natural killer cell activity. I. A novel non-radioactive method based on time-resolved fluorescence. J. Immunol. Methods 86: 225-229. [Medline]
  15. Lasagni, L., M. Francalanci, F. Annunziato, E. Lazzeri, S. Giannini, L. Cosmi, C. Sagrinati, B. Mazzinghi, C. Orlando, E. Maggi, et al 2003. An alternatively spliced variant of CXCR3 mediates the inhibition of endothelial cell growth induced by IP-10, Mig, and I-TAC, and acts as functional receptor for platelet factor 4. J. Exp. Med. 197: 1537-1549. [Abstract/Free Full Text]
  16. Charbonnier, A. S., N. Kohrgruber, E. Kriehuber, G. Stingl, A. Rot, D. Maurer. 1999. Macrophage inflammatory protein 3a is involved in the constitutive trafficking of epidermal Langerhans cells. J. Exp. Med. 190: 1755-1768. [Abstract/Free Full Text]
  17. Lechleitner, S., R. Kunstfeld, C. Messeritsch-Fanta, K. Wolff, P. Petzelbauer. 1999. Peripheral lymph node addressins are expressed on skin endothelial cells. J. Invest. Dermatol. 113: 410-414. [Medline]
  18. Lechleitner, S., J. Gille, D. R. Johnson, P. Petzelbauer. 1998. Interferon enhances tumor necrosis factor-induced vascular cell adhesion molecule 1 (CD106) expression in human endothelial cells by an interferon-related factor 1-dependent pathway. J. Exp. Med. 187: 2023-2030. [Abstract/Free Full Text]
  19. Mohan, K., E. Cordeiro, M. Vaci, C. McMaster, T. B. Issekutz. 2005. CXCR3 is required for migration to dermal inflammation by normal and in vivo activated T cells: differential requirements by CD4 and CD8 memory subsets. Eur. J. Immunol. 35: 1702-1711. [Medline]
  20. Christensen, J. E., C. de Lemos, T. Moos, J. P. Christensen, A. R. Thomsen. 2006. CXCL10 is the key ligand for CXCR3 on CD8+ effector T cells involved in immune surveillance of the lymphocytic choriomeningitis virus-infected central nervous system. J. Immunol. 176: 4235-4243. [Abstract/Free Full Text]
  21. Hamann, D., P. A. Baars, M. H. Rep, B. Hooibrink, S. R. Kerkhof-Garde, M. R. Klein, R. A. van Lier. 1997. Phenotypic and functional separation of memory and effector human CD8+ T cells. J. Exp. Med. 186: 1407-1418. [Abstract/Free Full Text]
  22. Ehlert, J. E., C. A. Addison, M. D. Burdick, S. L. Kunkel, R. M. Strieter. 2004. Identification and partial characterization of a variant of human CXCR3 generated by posttranscriptional exon skipping. J. Immunol. 173: 6234-6240. [Abstract/Free Full Text]
  23. Sauty, A., R. A. Colvin, L. Wagner, S. Rochat, F. Spertini, A. D. Luster. 2001. CXCR3 internalization following T cell-endothelial cell contact: preferential role of IFN-inducible T cell {alpha} chemoattractant (CXCL11). J. Immunol. 167: 7084-7093. [Abstract/Free Full Text]
  24. Fukada, K., Y. Sobao, H. Tomiyama, S. Oka, M. Takiguchi. 2002. Functional expression of the chemokine receptor CCR5 on virus epitope-specific memory and effector CD8+ T cells. J. Immunol. 168: 2225-2232. [Abstract/Free Full Text]
  25. Tomiyama, H., T. Matsuda, M. Takiguchi. 2002. Differentiation of human CD8+ T cells from memory to memory/effector phenotype. J. Immunol. 168: 5538-5550. [Abstract/Free Full Text]
  26. Neel, N. F., E. Schutyser, J. Sai, G. H. Fan, A. Richmond. 2005. Chemokine receptor internalization and intracellular trafficking. Cytokine Growth Factor Rev. 16: 637-658. [Medline]
  27. Hess, C., T. K. Means, P. Autissier, T. Woodberry, M. Altfeld, M. M. Addo, N. Frahm, C. Brander, B. D. Walker, A. D. Luster. 2004. IL-8 responsiveness defines a subset of CD8 T cells poised to kill. Blood 104: 3463-3471. [Abstract/Free Full Text]
  28. Axelrod, P. I., B. Lorber, E. C. Vonderheid. 1992. Infections complicating mycosis fungoides and Sézary syndrome. J. Am. Med. Assoc. 267: 1354-1358. [Abstract/Free Full Text]
  29. Padovan, E., G. C. Spagnoli, M. Ferrantini, M. Heberer. 2002. IFN-{alpha}2a induces IP-10/CXCL10 and MIG/CXCL9 production in monocyte-derived dendritic cells and enhances their capacity to attract and stimulate CD8+ effector T cells. J. Leukocyte Biol. 71: 669-676. [Abstract/Free Full Text]
  30. Kohrgruber, N., M. Gröger, P. Meraner, E. Kriehuber, P. Petzelbauer, S. Brandt, G. Stingl, A. Rot, D. Maurer. 2004. Plasmacytoid dendritic cell recruitment by immobilized CXCR3 ligands. J. Immunol. 173: 6592-6602. [Abstract/Free Full Text]
  31. Ludwig, A., F. Schiemann, R. Mentlein, B. Lindner, E. Brandt. 2002. Dipeptidyl peptidase IV (CD26) on T cells cleaves the CXC chemokine CXCL11 (I-TAC) and abolishes the stimulating but not the desensitizing potential of the chemokine. J. Leukocyte Biol. 72: 183-191. [Abstract/Free Full Text]
  32. Mullins, I. M., C. L. Slingluff, J. K. Lee, C. F. Garbee, J. Shu, S. G. Anderson, M. E. Mayer, W. A. Knaus, D. W. Mullins. 2004. CXC chemokine receptor 3 expression by activated CD8+ T cells is associated with survival in melanoma patients with stage III disease. Cancer Res. 64: 7697-7701. [Abstract/Free Full Text]

Related articles in The JI:

IN THIS ISSUE

The JI 2007 179: 3389-3390. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Winter, D.
Right arrow Articles by Maurer, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Winter, D.
Right arrow Articles by Maurer, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS