CD4+CXCR4highCD69+ T Cells Accumulate in Lung Adenocarcinoma Ori Wald, Uzi Izhar, Gail Amir, Shani Avniel, Yochai Bar-Shavit, Hanna Wald, Ido D. Weiss, Eithan Galun and This information is current as Amnon Peled of September 29, 2021. J Immunol 2006; 177:6983-6990; ; doi: 10.4049/jimmunol.177.10.6983 http://www.jimmunol.org/content/177/10/6983 Downloaded from

References This article cites 49 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/177/10/6983.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 29, 2021 *average

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

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

CD4؉CXCR4highCD69؉ T Cells Accumulate in Lung Adenocarcinoma

Ori Wald,1* Uzi Izhar,1*‡ Gail Amir,† Shani Avniel,* Yochai Bar-Shavit,* Hanna Wald,* Ido D. Weiss,* Eithan Galun,* and Amnon Peled2*

The chemokine receptor CXCR4 is involved in the growth and metastasis of tumor cells. However, the expression of its ligand, the chemokine CXCL12, in tumors and its role in regulating the accumulation of immune cells within the tumors is not clear. Using ELISA and immunohistochemistry we found that CXCL12 is expressed in the majority of nonsmall cell lung cancer tissue sections obtained from stage IA to IIB nonsmall cell lung cancer patients undergoing operation. Histopathologic examination of these sections indicated that high CXCL12 expression correlated with increased tumor inflammation. In addition, disease recurrence rates in a subgroup of adenocarcinoma patients showed a tendency to correlate with high

CXCL12 expression in the tumor. Isolation of adenocarcinoma-infiltrating immune cells demonstrated an increase in the Downloaded from percentage of CD4؉CD69؉CXCR4؉ T cells as compared with normal lung tissue. About 30% of these cells expressed the regulatory markers CD25high and FoxP3. The percentage of CD8 T cells within the tumor did not change, however; the percentage of NK and NK T cells was significantly reduced. In correlation with CXCR4 expression, CD4 T cells showed increased migration in response to CXCL12 compared with CD8 T cells and NK cells. Overall, these observations suggest that CXCL12 expression may influence tumor progression by shaping the immune cell population infiltrating lung adeno- carcinoma tumors. The Journal of Immunology, 2006, 177: 6983–6990. http://www.jimmunol.org/

ung cancer is the leading cause of cancer death in the phages into tumors (10). These cells in turn, support tumor growth United States and throughout the world. The pathogenesis by the secretion of various factors, among them the immune-mod- L of nonsmall cell lung cancer (NSCLC)3 is critically linked ulating TGF-␤ and the proangiogenic chemokine CXCL8 to cigarette smoking and exposure to secondhand smoke (1). It is (11–14). TGF-␤, among other immune-modulating , is estimated that in 2006 ϳ174,000 new cases of lung cancer will be also produced by regulatory T cells. It was shown that the bone diagnosed and that ϳ162,000 cases of death will result from lung marrow is a reservoir of regulatory T cells that traffic to this site cancer. NSCLC accounts for nearly 80% of these cases (1). through CXCL12/CXCR4 interactions (15). The trafficking of reg- by guest on September 29, 2021 Chemokines, a subfamily of chemotactic cytokines, play a cen- ulatory T cells into NSCLC tissue has not been studied to date. tral role in regulating immune cell trafficking in the body (2, 3). Chemokine receptor expression is a critical step in tumor me- Furthermore, a complex network of chemokines/chemokine recep- tastasis. The most studied chemokine receptor in this regard is tor interactions is involved in the development of primary malig- CXCR4, which is expressed by 23 different cancerous tissues (2, nant tumors and in the process of metastasis (2, 3). Several such 16, 17). Recent reports indicated that CXCR4, one of the two interactions participate in the development, progression, and me- known receptors for CXCL12, is expressed in NSCLC cells (18, tastasis of NSCLC. Angiogenesis in NSCLC is mediated in part by 19). It has been suggested that CXCL12/CXCR4 interactions fa- the chemokines CXCL8 and CXCL5, and their expression corre- cilitate the metastasis of the NSCLC tumor to the bone marrow and lates with tumor vascularity (4–6). Moreover, high CXCL8 dissemination into the pleural cavity (18, 20). Nonetheless, mRNA levels in stage I NSCLC were associated with shortened CXCL12 is highly expressed in the normal lung and is associated survival (7). Chemokines also regulate the recruitment of subpopu- with the establishment of metastasis of breast and colon carcinoma lations of immune cells into tumors (8). For example, CCL4, in the lung (21, 22). The expression of CXCL12 in NSCLC cells CCL5, and CXCL10 expression in lung adenocarcinoma was as- has not been clearly characterized to date. Phillips et al. (18) sociated with the presence of an active lymphocytic response in the reported that NSCLC cell lines (A549 and Calu-1) do not ex- tumor (9). Similarly, CCL2 produced by NCSLC recruits macro- press CXCL12 and suggested that CXCL12 is not expressed in lung adenocarcinoma tumors. Using immunohistochemistry and ELISA, we analyzed CXCL12 expression in patients with stage † ‡ *Goldyne Savad Institute of Therapy, Department of Pathology, and Depart- IA-IIB NSCLC. CXCL12 staining was detected in 42 of 46 tumor ment of Cardiothoracic Surgery, Hadassah University Hospital, Jerusalem, Israel samples tested and confirmed by ELISA. Isolated lung adenocar- Received for publication February 27, 2006. Accepted for publication September 7, 2006. cinoma and normal lung infiltrating were immuno- The costs of publication of this article were defrayed in part by the payment of page stained for CD3, CD4, CD8, CD25, FoxP3, CD56, CD69, and the charges. This article must therefore be hereby marked advertisement in accordance CXCL12 receptor CXCR4. In lung adenocarcinoma, relative to the with 18 U.S.C. Section 1734 solely to indicate this fact. normal lung decreased percentages of NK and NK T cells (12 vs 1 O.W. and U.I. contributed equally to this work. 28.8%, p Ͻ 0.01 and 6.4 vs 13.8%, p Ͻ 0.05, respectively) and 2 Address correspondence and reprint requests to Dr. Amnon Peled, Hadassah Uni- increased percentages of CD4ϩ T cells, (48.8 vs 24.5%, p Ͻ 0.01) versity Hospital, Gene Therapy Institute, P.O. Box 12000, Jerusalem, Israel. E-mail ϩ address: [email protected] were found. Increased percentages of tumor infiltrating CD4 T cells Ͻ 3 Abbreviations used in this paper: NSCLC, nonsmall cell lung cancer; HIF, hypoxia- expressed the early activation marker CD69 (66.1 vs 41.8%, p inducible factor; PTC, papillary thyroid carcinoma. 0.01), the marker CD25high (18 vs 7.6%, p Ͻ 0.01),

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 6984 CXCL12 ROLE IN LUNG ADENOCARCINOMA IMMUNE CELL INFILTRATION and CXCR4 (62.8 vs 49.5%, p Ͻ 0.01). Furthermore, lung adenocar- The most common histologic type of lung cancer diagnosed in Israel and cinoma-derived CD4 T cells migrated better than CD8 T cells and NK in the western world is of the adenocarcinoma type (23). Thirty samples of cells in response to CXCL12. CXCL12/CXCR4 interactions may fresh lung adenocarcinoma tumor tissue and adjacent tumor-free lung tis- ϩ sue removed during operation were collected. In addition, fresh blood sam- therefore play a role in the accumulation CXCR4 T cells and a ples from these patients were obtained at the time of operation. Lympho- high subset of CD25 regulatory T cells within NSCLC tissue. cytes were extracted from adenocarcinoma tumor tissue and adjacent tumor-free normal lung tissue for phenotypic characterization and migra- Materials and Methods tion assays. In addition, ELISA was used to assess the expression of CXCL12 in homogenates derived from these tissues and in the peripheral Tissue collection, histological evaluation, and patient-specific blood. Protocols for these experiments are described below. clinical data Immunohistochemistry Fresh human lung specimens, paraffin-embedded tissue sections, and fresh blood samples were obtained according to the guidelines approved by the NSCLC tissue samples were routinely fixed with formalin and embedded Hadassah Hospital Ethics Committee (Jerusalem, Israel). Forty-nine pa- in paraffin. Ag retrieval was performed in EDTA buffer for 15 min in a tients with early stage NSCLC (clinical stage IA-IIB) that underwent com- microwave, and sections were stained with the mAb 79018 (R&D Systems) plete tumor resection in our institute between January 2000 to October for CXCL12 (1/100), using a standard indirect avidin-biotin HRP method 2005 participated in the study. No chemotherapy or radiotherapy was ad- according to the manufacturer’s instructions. 3-Amino-9-ethylcarbazole ministered before the operation and the collection of tissue samples to was used for color development and sections were counterstained with exclude the effects of such treatments on the results. Histological sections hematoxylin. As negative controls, sections were stained either with no were prepared from formalin-fixed, paraffin-embedded tissues and stained primary Ab (PBS) or with an isotype-matched control Ab (24, 25). with H&E. Histopathological diagnosis was confirmed for each specimen. ELISA

Postoperative staging indicated that three of the 49 patients had a higher Downloaded from stage of disease; these patients were omitted from the study. Of the 46 lung Tumor tissue homogenate and normal lung tissue homogenate were pre- tumor samples, 29 samples were of lung adenocarcinoma, and 17 samples pared on ice from weighed tissue samples. ELISA for CXCL12 in the were of squamous cell carcinoma. Analysis of CXCL12 expression by patients plasma, normal lung tissue homogenate, and the NSCLC tissue immunohistochemistry was performed on these tumors. Sections were homogenate were performed using the Quantikine (R&D Systems) ac- blindly assessed by a pathologist and scored for the extent and intensity of cording to the manufacturer’s instructions. CXCL12 staining; in addition, tumor inflammation was scored. Intensity of staining was scored between 0 and 3. The extent of staining, i.e. the per- Isolation of normal lung and tumor tissue-derived lymphocytes centage of tumor cells expressing CXCL12, was scored between 1 and 4 http://www.jimmunol.org/ (score of 1: 0–25%; score of 2: 25–50%; score of 3: 50–75%; and score of Normal lung and adenocarcinoma tumor tissue obtained during operations 4: 75–100% of tumor cells that stained positively to CXCL12). Inflamma- were washed twice in 10 ml of PBS to remove the remaining blood. Tissue ϳ tion was scored between 0 and 3; lymphocytes were used as indicators of was cut into 1-mm pieces and incubated in 2 ml of RPMI 1640 (Invitro- inflammation. A score of 0 was assigned to slides where no gen Life Technologies) for 20 min with collagenase type IV (Worthington collection were identified; a score of 1 was assigned to slides with only few Biochemical) and DNase I (Worthington Biochemical) at 37°C. Following small lymphocyte collections; a score of 2 was assigned to slides were incubation, the tissue was run through a 1-ml pipette and laid for separation multiple small collections of lymphocytes were identified; and a score 3 over Ficoll/Histopaque gradients (Sigma-Aldrich). Cells were washed and was assigned to slides with multiple large lymphocyte collections. Scoring centrifuged at 1,200 rpm for 5 min at 4°C. Pellets were resuspended in of inflammation was done blindly. The sum of the scores for the extent and FACS buffer for flow cytometry analysis and in a migration medium for intensity of CXCL12 staining was calculated and correlated to tumor in- migration assays. Patients with early disease stages IA to IIB usually have by guest on September 29, 2021 flammation. Furthermore, this score was examined in regard to long-term small to moderate tumors. Furthermore, only a small section from the core disease progression. Fourteen of 46 patients participating in the study had of the tumor may be excised for scientific analysis, as the margins must be recurrence of disease. Parameters regarding patients participating in the examined to assure complete resection. Often, small numbers of total tu- ϫ 5 ϫ 6 study are presented in Table I. mor cells between 1 10 to 1 10 cells per biopsy ware obtained, limiting the extent of experiments performed.

Table I. Patient characteristics and tumor CXCL12 staininga Phenotypic analysis of lymphocytes and flow cytometry analysis Isolated lymphocytes were taken at a concentration of 1–5 ϫ 105 cells/ml Values in 0.1 ml of PBS containing 0.1% BSA (Biological Industries) and 0.01% sodium azide (FACS buffer). Fc receptors were blocked by incubation with Patients 49 1% human plasma for 15 min on ice. Blocked cells were then mixed with Enrolled 46 FITC-, PE-, Cy3-, or allophycocyanin-conjugated CD3, CD4, CD8, (1/40; Omitted (postoperative stage, 3 IQ Products), CXCR4 (1/40; R&D Systems), CD25 (Miltenyi Biotec), IIB) CD69, and CD56 (BD Biosciences) mAbs or isotype controls for 20 min Male/female 26/20 on ice and washed with FACS buffer. A FoxP3 internal staining kit was Mean age Ϯ SD (years) 70 Ϯ 10.2 purchased from R&D Systems, and staining was performed according to Mean followup (years) 3.7 Ϯ 1.3 manufacturer protocol. Immunostained cells were analyzed by flow cytom- Pathology etry using the FACSCaliber flow cytometer (BD Biosciences); the data Adenocarcinoma 29 were analyzed using the CellQuest 3.3 software (BD Biosciences). For all Squamous cell carcinoma 17 analyses, lymphocyte gating was performed using the forward scatter vs Stage IA to IIB side scatter plot. Population of T cells (CD3ϩCD56Ϫ), NK T cells Recurrence (CD3ϩCD56ϩ), and NK cells (CD3ϪCD56ϩ) were defined. T cells were Adenocarcinoma recurrence 7 of 29 (24.1%) further classified according to CD4 and CD8 expression. Squamous recurrence 7 of 17 (41.2%) CXCL12 expression Transwell migration assays CXCL12 positive 42 or 46 (91.3%) Tumor-derived and normal lung-derived lymphocyte migration was as- Adenocarcinoma CXCL12 8 (none with recurrence; 0%) sessed in 24-well chemotaxis chambers (6.5-mm diameter, 5-mm pore (scores of 0–3) polycarbonate Transwell culture insert; Costar). Six hundred microliters of Adenocarcinoma CXCL12 21 (7 with recurrence; 33.3%) RPMI 1640 with 1% BSA (migration buffer) with or without 250 ng/ml (scores of 4–7) CXCL12␣ (R&D Systems) was added to the lower wells. Cells (1 ϫ 105) Squamous CA CXCL12 4 (2 with recurrence; 50%) suspended in 100 ␮l of RPMI 1640 with 1% BSA were added to the upper (scores of 0–3) wells. Each experiment was performed in triplicate. After incubation for Squamous CA CXCL12 13 (5 with recurrence; 38.5%) 3 h, the membrane was removed and migrating cells were collected from (scores of 4–7) the lower well and immunostained for the cell markers CD3, CD4, CD8, a The characteristics of patients participating in the study are presented. The cor- and CD56. These cells were analyzed by flow cytometry. Each sample was relation between the scores of CXCL12 staining in patients with adenocarcinoma and acquired for 1 min, and the number of each subpopulation of migrating squamous cell carcinoma (CA) and disease recurrence is shown. cells was determined. For comparison between different experiments, the The Journal of Immunology 6985

FIGURE 1. Isolation and characterization of lung adenocarcinoma resident immune cells. Representa- tive FACS analysis (A) and average percentage (B)of NK, NK T, CD4, and CD8 T cells in adenocarcinoma tissue (Tumor) and adjacent tumor-free lung tissue Downloaded from .p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .(Normal) are shown (n ϭ 22) http://www.jimmunol.org/ by guest on September 29, 2021

percentage of CD4 and CD8 T cells and NK cells migrating in response to terization of the subtype of CD4 T cells accumulating in the tu- CXCL12␣ was calculated. The percentage of migrating cells for each sub- mors was done by staining these cells for the markers CD45RO, population of immune cells was determined as follows. 1) The number of CD45RA, CD69, CD25, and the transcription factor FoxP3. Nine- cells used at the beginning of the experiment and laid into the upper well, ϩ representing 100% of the cells, was determined by acquisition for 1 min ty-five percent of tumor CD4 T cells express CD45RO, which is and was termed the total. 2) The number of cells in the lower well, where a marker for activated and memory T cells as compared with 75% no chemokine was added, was determined after3hofincubation and was in the normal lung (Fig. 2A). A significant increase in the percent- termed spontaneous migration. 3) The number of cells in the lower well, age of CD4ϩ T cells expressing the early activation marker CD69 where CXCL12␣ was present, was determined after3hofincubation and Ͻ ϭ was termed migration. The percentage of chemokine-induced migration was also observed in the tumor (66.1 vs 41.8%, p 0.01; n 18) was then calculated as follows: percentage of migration ϭ (migration Ϫ (Fig. 2B). Regulatory T cell may play a critical role in modulating spontaneous migration)/total ϫ 100. the immune response against tumors. We found an increased per- centage of CD4ϩ T cells expressing the regulatory markers Results CD25high and FoxP3ϩ in the tumor compared with the normal lung ϩ CD4 T cells accumulate in lung adenocarcinoma (18 vs 7.6%, p Ͻ 0.01; n ϭ 17 and 29.1 vs 11.7%, p Ͻ 0.01; n ϭ Adenocarcinoma and normal lung resident lymphocytes were iso- 8 respectively) (Fig. 2, B and D). Interestingly we did not find lated and analyzed for expression of the cell markers for CD40L expression on these cells (data not shown). Overexpression ϩ Ϫ ϩ ϩ ϩ ϩ ␤ CD56 CD3 NK cells, CD3 CD56 NK T cells, CD3 CD4 T of CD69 was shown to induce the secretion of TGF- from CD4 T ϩ high ϩ cells, and CD3ϩCD8ϩ T cells (Fig. 1A). The relative percentage of cells (26, 27); furthermore, CD4 CD25 FoxP3 regulatory T cells ␤ CD4 T cells, CD8 T cells, NK cells, and NK T cells from total were also shown to secrete high levels of TGF- (28). The accumu- lymphocytes were calculated (Fig. 1B). We found that the percent- lation of these cells in adenocarcinoma lung tumors may therefore age of NK cells and NK T cells was significantly reduced (12 vs contribute to suppression of the immune response against this tumor. 28.8%, p Ͻ 0.01 and 6.4 vs 13.8%, p Ͻ 0.05, respectively; n ϭ 22) in tumor tissue relative to that for the normal lung (Fig. 1B). Fur- CXCR4/CXCL12 interactions are involved in the recruitment/ thermore, CD8 T cells failed to accumulate in tumor tissue retention of CD4 T cells in lung adenocarcinoma whereas the percentage of CD4 T cell significantly increased (48.8 CXCR4 was shown to play a role in the trafficking and tissue vs 24.5%, p Ͻ 0.01; n ϭ 22) in tumor tissue (Fig. 1B). Our findings localization of CD4ϩ T cells and regulatory T cells (15, 29). To indicate that CD4 T cells specifically accumulate in adenocarci- study whether CXCR4 is involved in the trafficking and tissue noma relative to adjacent tumor-free lung tissue. Further charac- localization of CD4 T cells in the lung and in adenocarcinoma 6986 CXCL12 ROLE IN LUNG ADENOCARCINOMA IMMUNE CELL INFILTRATION Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021 FIGURE 2. RO, RA, CD69, CD25, and FoxP3 expression in tumor-infiltrating CD4 T cells. RO/RA (A), CD69 (n ϭ 18) (B), CD25 (n ϭ 17) (C), and FoxP3 (n ϭ 8) (D) expression in lung adenocarcinoma (Tumor) CD4 T cells and adjacent tumor-free lung (Normal) CD4 T cells was determined. Representative FACS analysis and the percentage of positive cells in each patient are shown, and the average percentage of positive cells for each marker .(p Ͻ 0.01 ,ءء) .(is also shown. For RO and RA expression, mean percentage of positive cells is shown (n ϭ 5 tumors we analyzed the expression of CXCR4 by CD4 T cells, derived CD8 T cells to respond to CXCL12 was significantly re- CD8 T cells, and NK cells derived from these tissues. In all sam- duced as compared with tumor-derived CD4 T cells (Fig. 4B). In ples tested, considerable percentages of tumor- and normal lung- normal lung tissue CXCL12 is expressed by epithelial cells and is derived CD4 T cells and CD8 T cells expressed CXCR4, whereas involved in the recruitment of immune cell under normal and tumor- and normal lung-derived NK cells expressed only marginal pathological conditions such as asthma (30). The expression of levels of CXCR4 (Fig. 3A). Furthermore, we found that in both the CXCL12 in NSCLC cells has not yet been tested. Using ELISA, lung and the tumor a higher percentage of CD4ϩ T cells stained we found that the levels of CXCL12 in a normal lung and in positive for CXCR4 (49.5 and 62.8%, respectively; n ϭ 22) as adenocarcinoma tumors is 5-fold higher compared with the levels compared with CD8ϩ T cells (31.7 and 47.8%, respectively; n ϭ of CXCL12 in the serum (Fig. 5A). Immunohistochemistry anal- 9) and NK cells (19.7 and 23.5%, respectively; n ϭ 9) (Fig. 3, ysis confirmed that in NSCLC tumor cells expressed CXCL12 in A–D). In addition, in the tumor the level of CXCR4 expression 42 of 46 (91.3%) cases tested (Table I and Fig. 5, B and C). significantly increased on both CD4 and CD8 T cells as compared CXCL12 levels of expression by tumor cells correlated with the with NK cells. Furthermore, in most cases tested the percentages level of tumor inflammation in a subgroup of adenocarcinoma but and levels of CXCR4 on CD4 and CD8 T cells but not NK cells not squamous carcinoma patients (Fig. 5D). Representative stain- was increased in the tumor relative to tumor adjacent normal lung ing of a tumor that expresses a high level of CXCL12 and is highly tissue. Moreover, most of the CD4ϩ T cells that express FoxP3 inflamed is shown in Fig. 5B. Representative staining of a tumor also express CXCR4 (Fig. 3E). We next tested the migratory re- that expresses a low level of CXCL12 and is not inflamed is shown sponse of tumor- and normal lung-derived CD4 and CD8 T cells in Fig. 5C. Overall, these results suggest a high degree of corre- and NK cells in response to CXCL12. Because the yield of tumor lation between CXCL12 expression and degree of inflammation NK cell was very low, we compared the migration of CD4 and and the accumulation of CD4 T cells in adenocarcinoma. In this CD8 cells from a normal lung to that of normal lung NK cells. We study group, adenocarcinoma patients with a low score of found that both CD4 and CD8 T cells migrate better than NK cells CXCL12 staining did not show disease recurrence (0 of 8). The in response to CXCL12 (Fig. 4A). CD4 and CD8 T cells derived recurrence of disease in adenocarcinoma patients with high score from fresh adenocarcinoma samples also migrate well in response of CXCL12 staining was higher (7 of 21). In squamous cell carci- to CXCL12. In three of four cases studied, the ability of tumor- noma no difference between the two groups were found (Table I). The Journal of Immunology 6987 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 3. CXCR4 expression in lung adenocarcinoma and adjacent tumor-free, lung tissue-resident lymphocytes. A, The mean percentage of CXCR4 expression in CD4 and CD8 T cells and NK cells (n ϭ 22, 9, and 9 respectively) derived from lung adenocarcinoma (Tumor) and adjacent tumor-free lung tissue (Normal) was determined. B–D, Representative FACS analysis and the comparative percentage of CXCR4 positive cells in each patient are shown for CD4 T cells (B) CD8 T cells (C), and NK cells (D)(n ϭ 9). E, A representative double staining of CD4 T cells derived from lung adenocarcinoma .p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .(and adjacent tumor-free lung tissue for FoxP3 and CXCR4 is presented (n ϭ 3

Taken together, these observations indicate a tendency of diseases pression is positively regulated by the human epithelial growth recurrence to correlate with high CXCL12 expression in the tumor. factor receptor 2 (HER2) (33) and that neutralizing the interaction However, the numbers of participants in this study is not sufficient for between CXCL12 and CXCR4 by either anti-CXCR4 Abs or by statistical significance to be obtained. Larger studies are therefore re- silencing the expression of CXCR4 using RNA interference tech- quired; conducting such a study is beyond the scope of this report. nology significantly impaired metastasis and progression of the tumor in vivo (21, 34). In regard to prostate cancerous tumors it Discussion was shown that overexpression of CXCR4 in the tumors is associ- CXCR4 was found to play a critical role in the progression and ated with increased tumor growth, angiogenesis, and metastasis (31). development of various tumors including breast, prostate, and Schioppa et al. (35) have shown that CXCR4 is induced by hypoxic clear cell renal carcinoma (21, 31, 32). In clear cell renal carci- conditions via the activation and transcript stabilization of the noma, CXCR4 expression correlates with increased tumor metas- transcriptional activator hypoxia-inducible factor (HIF) 1. This tasis (32). In breast cancer tumors it was shown that CXCR4 ex- report has further suggested that hypoxia may increase CXCR4 6988 CXCL12 ROLE IN LUNG ADENOCARCINOMA IMMUNE CELL INFILTRATION

FIGURE 4. Migration of tumor and adjacent tumor-free lung tissue CD4 and CD8 T cells and NK cells in response to CXCL12. A, Because the number of NK cells in lung adenocarcinoma tu- mors is very low, we compared the migration of normal lung NK cells to that of normal lung CD4 and CD8 T cells in response to 250 ng/ml CXCL12 (n ϭ 2). B, The percentage of migration of CD4 and CD8 T cells derived from lung ade- nocarcinoma in response to 250 ng/ml CXCL12 .p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .(was measured (n ϭ 4 Downloaded from http://www.jimmunol.org/

expression and thus promote the metastatic potential of cancer and CXCL12 may promote tumor growth and invasiveness (31). cells (35). Indeed, it was recently shown that HIF-1 regulates The RET/PTC1 oncogene is expressed in up to 30% of papillary CXCR4 expression in NSCLC cells and may thus increase their thyroid carcinoma (PTC) tumors (43). It was recently reported that metastatic potential (36, 37). Further support for this idea is pro- the RET/PTC1 oncogene induces the up-regulation of several in- vided by findings indicating that, in NSCLC tissue, HIF-1 is flammatory in primary human thyrocytes, and among these mainly expressed by squamous cell carcinoma at necrotic areas genes were both CXCR4 and CXCL12 (43, 44). The authors sug- by guest on September 29, 2021 (38, 39) and that HIF-1 expression is associated with a worse prog- gested that CXCL12 may act in an autocrine fashion on PTC cells nosis (40, 41). Additional reports have also indicated that CXCR4 to induce their proliferation and survival (43). Similarly it was is expressed on NSCLC cells (18–20, 37). It was further shown shown that CXCR4/CXCL12 interactions promote breast carci- that NSCLC cell lines express CXCR4 but not CXCL12 (18). By noma tumor growth by inducing tumor cell proliferation and an- using NSCLC cell lines and anti-CXCR4 Abs it was demonstrated giogenesis (45). It is possible that CXCL12/CXCR4 interaction that CXCR4/CXCL12 interactions facilitate metastasis of NSCLC within the NSCLC microenvironment delivers antiapoptotic/pro- cell lines to the bone marrow (18). Oonakahara et al. (20) have liferative signals to tumor cells and promotes disease progression. shown that mesothelial cell produce high levels of CXCL12 and Moreover, CXCL12 in combination with additional proangiogenic suggested that the dissemination of NSCLC cells into the pleural factors produced at the tumor microenvironment may favor angio- cavity is mediated by CXCR4. Taken together, these data may genesis and increase tumor vascularity. suggest that hypoxia-induced CXCR4 expression by NSCLC tu- In addition to its role in stimulating tumor growth and angio- mor cells may promote their metastatic potential and dissemination genesis, CXCL12 may also participate in the recruitment and re- to the bone marrow and pleural cavity. tention of immune cells in the tumor (2, 16, 17). This possibility Interestingly HIF-1 was also shown to up-regulate the expres- may be supported by the fact that CXCL12 expression in NSCLC sion of CXCL12 in a manner similar to that of CXCR4 and to cells is significantly higher than in the serum. We have found that promote the recruitment of progenitor cells to hypoxic areas (42). CD4ϩ T cells specifically accumulate within lung adenocarcino- Using immunohistochemical staining and ELISA, we found that mas, whereas effector NK cells are depleted from the tumor mi- the majority of NSCLC tumors in vivo expresses CXCL12. How- croenvironment and CD8ϩ T cells are not enriched. Tumor-resi- ever, in our study the expression of CXCL12 was not restricted to dent CD4ϩ T cells were shown to express high levels of CXCR4 necrotic areas; nonetheless, CXCL12 was also detected at necrotic relative to NK and CD8 cells and to migrate better than these areas of some tumors. Furthermore, high expression of CXCL12 in subtypes of cells in response to CXCL12. Lung adenocarcinoma the tumors was found in 33% of the patients with recurrence but tissue homogenates had high levels of CXCL12 relative to the was not found in patients without recurrence. It is therefore pos- peripheral blood. Furthermore, high CXCL12 expression in lung sible that, similar to the correlation of HIF expression with NSCLC adenocarcinomas is associated with increased tumor inflammation. disease outcome, HIF may be involved in up-regulation of The accumulation of CD4ϩ T cells expressing high CXCR4 levels CXCL12 and thus support tumor progression. was shown in RA and in fibrotic liver where the levels of CXCL12 The expression of CXCL12 by the majority of NSCLC tumors expression are high (24, 29). Furthermore, overexpression of CXCR4 indicates that CXCR4-mediated metastasis of NSCLC tumor cells in CD4ϩ T cells induced their accumulation in tissues that express into the pleural cavity cannot be simply attributed to the migration high levels of CXCL12 such as the bone marrow (46). We therefore of tumor cells into CXCL12-expressing tissues and that CXCR4 suggest that CXCR4/CXCL12 interactions are involved in increasing The Journal of Immunology 6989 Downloaded from http://www.jimmunol.org/

FIGURE 5. NSCLC CXCL12 expression and correlation with tumor inflammation. A, ELISA was used to determine CXCL12 expression in lung by guest on September 29, 2021 adenocarcinoma (Tumor), adjacent tumor-free lung tissue (Normal), and the peripheral blood (Serum) of NSCLC patients (n ϭ 15). B and C, Representative immunohistochemical staining of CXCL12 in NSCLC. High and low CXCL12 expression in NSCLC is shown. D, The correlation of CXCL12 expression .p Ͻ 0.05 ,ء .with tumor inflammation as assessed blindly by a specialized histopathologist is shown

the recruitment/retention of CD4ϩ T cells in lung adenocarcinomas. may also be involved in their recruitment into adenocarcinomas (15). Nevertheless, it should be noted that NSCLC tumors were shown to Indeed, CD4ϩFoxP3ϩ regulatory T cells in the adenocarcinomas ex- expresses a wide array of chemokines such as CCL2, CCL4, CCL5, pressed CXCR4. CXCL1, CXCL3, CXCL5, CXCL8, and CXCL10 and that these che- Further immune modulation in adenocarcinoma may be therefore mokines may also influence the recruitment/retention of distinct sub- mediated by the migration of CD4ϩCD25high regulatory T cell into populations of immune cells to the tumor (9, 10, 47). the tumor and by the secretion of TGF-␤. In addition to modulation of The roles of CD4ϩ T cells in adenocarcinoma pathogenesis are not the immune response, TGF-␤ has been shown to up-regulate the ex- clear. We have found that tumor resident CD4ϩ T cells express high pression of CXCR4 on naive T cells and to increase their migration in levels of the early activation marker CD69 relative to normal lung response to CXCL12 (50). We therefore suggest that a cycle of tissue resident cells. Nevertheless, other activation markers such as TGF-␤ secretion leading to CXCR4 up-regulation and the retention of CD40L were not detected (data not shown). CD69 was recently char- additional CD3ϩCD4ϩCD69ϩ cells capable of releasing TGF-␤ in acterized as an immune-modulating molecule that is up-regulated fol- the tumor microenvironment may contribute to adenocarcinoma lowing cell activation and mediates the secretion of TGF-␤. CD69- pathogenesis. deficient mice showed reduced autoimmune reactivity and increased anti-tumor activity (26, 27). It is therefore possible that the presence Acknowledgments of CD4ϩ T cells that express high levels of CD69 is linked to TGF-␤ We thank Mary Clausen (Gene Therapy Institute, Hadassah University secretion and mediates tumor evasion of the immune system. Indeed, Hospital, Jerusalem, Israel) for technical assistance. ␤ immunostaining for TGF- in NSCLC tissue has shown that this cy- Disclosures tokine is mainly produced by tumor-infiltrating immune cells and is The authors have no financial conflict of interest. associated with a more aggressive pattern of disease (14). We found that CD4ϩCD25highFoxP3ϩ regulatory T cells are accumulating in References adenocarcinoma tumors. Similar findings regarding the accumulation 1. Rosell, R., E. Felip, R. Garcia-Campelo, and C. Balana. 2004. The biology of of CD4ϩCD25high T cells in NSCLC tissue were reported by Woo non-small-cell lung cancer: identifying new targets for rational therapy. Lung Cancer 46: 135–148. et al. (48, 49). CXCL12/CXCR4 interactions were shown to regulate 2. Zlotnik, A. 2004. Chemokines in neoplastic progression. Semin. Cancer Biol. 14: regulatory T cells as well as T cells trafficking to the bone marrow and 181–185. 6990 CXCL12 ROLE IN LUNG ADENOCARCINOMA IMMUNE CELL INFILTRATION

3. Homey, B., A. Muller, and A. Zlotnik. 2002. Chemokines: agents for the immu- suppressive activity: the role of IL-2, TGF-␤, and IL-10. J. Immunol. 172: notherapy of cancer? Nat. Rev. Immunol. 2: 175–184. 5213–5221. 4. Pold, M., L. X. Zhu, S. Sharma, M. D. Burdick, Y. Lin, P. P. Lee, A. Pold, J. Luo, 29. Nanki, T., K. Hayashida, H. S. El-Gabalawy, S. Suson, K. Shi, H. J. Girschick, K. Krysan, M. Dohadwala, et al. 2004. Cyclooxygenase-2-dependent expression S. Yavuz, and P. E. Lipsky. 2000. Stromal cell-derived factor-1-CXC chemokine of angiogenic CXC chemokines ENA-78/CXC Ligand (CXCL) 5 and interleukin- receptor 4 interactions play a central role in CD4ϩ T cell accumulation in rheu- 8/CXCL8 in human non-small cell lung cancer. Cancer Res. 64: 1853–1860. matoid arthritis synovium. J. Immunol. 165: 6590–6598. 5. Arenberg, D. A., M. P. Keane, B. DiGiovine, S. L. Kunkel, S. B. Morris, 30. Lukacs, N. W., A. Berlin, D. Schols, R. T. Skerlj, and G. J. Bridger. 2002. Y. Y. Xue, M. D. Burdick, M. C. Glass, M. D. Iannettoni, and R. M. Strieter. AMD3100, a CxCR4 antagonist, attenuates allergic lung inflammation and air- 1998. Epithelial-neutrophil activating peptide (ENA-78) is an important angio- way hyperreactivity. Am. J. Pathol. 160: 1353–1360. genic factor in non-small cell lung cancer. J. Clin. Invest. 102: 465–472. 31. Darash-Yahana, M., E. Pikarsky, R. Abramovitch, E. Zeira, B. Pal, R. Karplus, 6. Arenberg, D. A., S. L. Kunkel, P. J. Polverini, M. Glass, M. D. Burdick, and K. Beider, S. Avniel, S. Kasem, E. Galun, and A. Peled. 2004. Role of high R. M. Strieter. 1996. Inhibition of interleukin-8 reduces tumorigenesis of human expression levels of CXCR4 in tumor growth, vascularization, and metastasis. non-small cell lung cancer in SCID mice. J. Clin. Invest. 97: 2792–2802. FASEB J. 18: 1240–1242. 7. Yuan, A., P. C. Yang, C. J. Yu, W. J. Chen, F. Y. Lin, S. H. Kuo, and K. T. Luh. 32. Staller, P., J. Sulitkova, J. Lisztwan, H. Moch, E. J. Oakeley, and W. Krek. 2003. 2000. Interleukin-8 messenger ribonucleic acid expression correlates with tumor Chemokine receptor CXCR4 downregulated by von Hippel-Lindau tumour sup- progression, tumor angiogenesis, patient survival, and timing of relapse in non- pressor pVHL. Nature 425: 307–311. small-cell lung cancer. Am. J. Respir. Crit. Care Med. 162: 1957–1963. 33. Li, Y. M., Y. Pan, Y. Wei, X. Cheng, B. P. Zhou, M. Tan, X. Zhou, W. Xia, 8. Mantovani, A., P. Allavena, S. Sozzani, A. Vecchi, M. Locati, and A. Sica. 2004. G. N. Hortobagyi, D. Yu, and M. C. Hung. 2004. Upregulation of CXCR4 is Chemokines in the recruitment and shaping of the leukocyte infiltrate of tumors. essential for HER2-mediated tumor metastasis. Cancer Cell 6: 459–469. Semin. Cancer Biol. 14: 155–160. 34. Liang, Z., Y. Yoon, J. Votaw, M. M. Goodman, L. Williams, and H. Shim. 2005. 9. Moran, C. J., D. A. Arenberg, C. C. Huang, T. J. Giordano, D. G. Thomas, Silencing of CXCR4 blocks breast cancer metastasis. Cancer Res. 65: 967–971. D. E. Misek, G. Chen, M. D. Iannettoni, M. B. Orringer, S. Hanash, and D. G. Beer. 2002. RANTES expression is a predictor of survival in stage I lung 35. Schioppa, T., B. Uranchimeg, A. Saccani, S. K. Biswas, A. Doni, A. Rapisarda, adenocarcinoma. Clin. Cancer Res. 8: 3803–3812. S. Bernasconi, S. Saccani, M. Nebuloni, L. Vago, et al. 2003. Regulation of the 10. Arenberg, D. A., M. P. Keane, B. DiGiovine, S. L. Kunkel, S. R. Strom, chemokine receptor CXCR4 by hypoxia. J. Exp. Med. 198: 1391–1402. M. D. Burdick, M. D. Iannettoni, and R. M. Strieter. 2000. Macrophage infiltra- 36. Phillips, R. J., J. Mestas, M. Gharaee-Kermani, M. D. Burdick, A. Sica, Downloaded from tion in human non-small-cell lung cancer: the role of CC chemokines. Cancer J. A. Belperio, M. P. Keane, and R. M. Strieter. 2005. Epidermal growth factor Immunol. Immunother. 49: 63–70. and hypoxia-induced expression of CXC chemokine receptor 4 on non-small cell 11. Mantovani, A., S. Sozzani, M. Locati, P. Allavena, and A. Sica. 2002. Macro- lung cancer cells is regulated by the phosphatidylinositol 3-kinase/PTEN/AKT/ phage polarization: tumor-associated macrophages as a paradigm for polarized mammalian target of rapamycin signaling pathway and activation of hypoxia ␣ M2 mononuclear phagocytes. Trends Immunol 23: 549–555. inducible factor-1 . J. Biol. Chem. 280: 22473–22481. 12. White, E. S., S. R. Strom, N. L. Wys, and D. A. Arenberg. 2001. Non-small cell 37. Phillips, R. J., J. Mestas, M. Gharaee-Kermani, M. D. Burdick, A. Sica, J. A. lung cancer cells induce monocytes to increase expression of angiogenic activity. Belperio, M. P. Keane, and R. M. Strieter. 2005. EGF and hypoxia-induced expres- sion of CXCR4 on non-small cell lung cancer cells is regulated by the PI3-kinase/ J. Immunol. 166: 7549–7555. http://www.jimmunol.org/ 13. Chen, J. J., P. L. Yao, A. Yuan, T. M. Hong, C. T. Shun, M. L. Kuo, Y. C. Lee, PTEN/AKT/mTOR signaling pathway and activation of HIF-1␣. J. Biol. Chem. and P. C. Yang. 2003. Up-regulation of tumor interleukin-8 expression by infil- 38. Lee, C. H., M. K. Lee, C. D. Kang, Y. D. Kim, D. Y. Park, J. Y. Kim, M. Y. Sol, and trating macrophages: its correlation with tumor angiogenesis and patient survival K. S. Suh. 2003. Differential expression of hypoxia inducible factor-1 ␣ and tumor in non-small cell lung cancer. Clin. Cancer Res. 9: 729–737. cell proliferation between squamous cell carcinomas and adenocarcinomas among 14. Saji, H., H. Nakamura, I. Awut, N. Kawasaki, M. Hagiwara, A. Ogata, operable non-small cell lung carcinomas. J. Korean Med. Sci. 18: 196–203. M. Hosaka, T. Saijo, Y. Kato, and H. Kato. 2003. Significance of expression of 39. Swinson, D. E., J. L. Jones, G. Cox, D. Richardson, A. L. Harris, and K. J. O’Byrne. TGF-␤ in pulmonary metastasis in non-small cell lung cancer tissues. Ann. Tho- 2004. Hypoxia-inducible factor-1 ␣ in non-small cell lung cancer: relation to growth rac Cardiovasc Surg. 9: 295–300. factor, protease and apoptosis pathways. Int. J. Cancer 111: 43–50. 15. Zou, L., B. Barnett, H. Safah, V. F. Larussa, M. Evdemon-Hogan, P. Mottram, 40. Kim, S. J., Z. N. Rabbani, M. W. Dewhirst, Z. Vujaskovic, R. T. Vollmer, S. Wei, O. David, T. J. Curiel, and W. Zou. 2004. Bone marrow is a reservoir for E. G. Schreiber, E. Oosterwijk, and M. J. Kelley. 2005. Expression of HIF-1␣, ϩ ϩ CD4 CD25 regulatory T cells that traffic through CXCL12/CXCR4 signals. CA IX, VEGF, and MMP-9 in surgically resected non-small cell lung cancer.

Cancer Res. 64: 8451–8455. Lung Cancer 49: 325–335. by guest on September 29, 2021 16. Balkwill, F. 2004. The significance of cancer cell expression of the chemokine 41. Giatromanolaki, A., M. I. Koukourakis, E. Sivridis, H. Turley, K. Talks, receptor CXCR4. Semin Cancer Biol. 14: 171–179. F. Pezzella, K. C. Gatter, and A. L. Harris. 2001. Relation of hypoxia inducible 17. Balkwill, F. 2004. Cancer and the chemokine network. Nat. Rev. Cancer 4: factor 1 ␣ and 2 ␣ in operable non-small cell lung cancer to angiogenic/molecular 540–550. profile of tumours and survival. Br. J. Cancer 85: 881–890. 18. Phillips, R. J., M. D. Burdick, M. Lutz, J. A. Belperio, M. P. Keane, and 42. Ceradini, D. J., A. R. Kulkarni, M. J. Callaghan, O. M. Tepper, N. Bastidas, R. M. Strieter. 2003. The stromal derived factor-1/CXCL12-CXC chemokine M. E. Kleinman, J. M. Capla, R. D. Galiano, J. P. Levine, and G. C. Gurtner. receptor 4 biological axis in non-small cell lung cancer metastases. 2004. Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 Am. J. Respir Crit. Care Med. 167: 1676–1686. induction of SDF-1. Nat. Med. 10: 858–864. 19. Spano, J. P., F. Andre, L. Morat, L. Sabatier, B. Besse, C. Combadiere, 43. Borrello, M. G., L. Alberti, A. Fischer, D. Degl’innocenti, C. Ferrario, P. Deterre, A. Martin, J. Azorin, D. Valeyre, et al. 2004. Chemokine receptor M. Gariboldi, F. Marchesi, P. Allavena, A. Greco, P. Collini, et al. 2005. Induc- CXCR4 and early-stage non-small cell lung cancer: pattern of expression and tion of a proinflammatory program in normal human thyrocytes by the RET/ correlation with outcome. Ann. Oncol. 15: 613–617. PTC1 oncogene. Proc. Natl. Acad. Sci. USA 102: 14825–14830. 20. Oonakahara, K., W. Matsuyama, I. Higashimoto, M. Kawabata, K. Arimura, and 44. Castellone, M. D., V. Guarino, V. De Falco, F. Carlomagno, F. Basolo, M. Osame. 2004. Stromal-derived factor-1␣/CXCL12-CXCR 4 axis is involved P. Faviana, M. Kruhoffer, T. Orntoft, J. P. Russell, J. L. Rothstein, et al. 2004. in the dissemination of NSCLC cells into pleural space. Am. J. Respir Cell Mol. Functional expression of the CXCR4 chemokine receptor is induced by RET/PTC Biol. 30: 671–677. oncogenes and is a common event in human papillary thyroid carcinomas. On- 21. Muller, A., B. Homey, H. Soto, N. Ge, D. Catron, M. E. Buchanan, cogene 23: 5958–5967. T. McClanahan, E. Murphy, W. Yuan, S. N. Wagner, et al. 2001. Involvement of 45. Orimo, A., P. B. Gupta, D. C. Sgroi, F. Arenzana-Seisdedos, T. Delaunay, chemokine receptors in breast cancer metastasis. Nature 410: 50–56. R. Naeem, V. J. Carey, A. L. Richardson, and R. A. Weinberg. 2005. Stromal 22. Zeelenberg, I. S., L. Ruuls-Van Stalle, and E. Roos. 2003. The chemokine re- fibroblasts present in invasive human breast carcinomas promote tumor growth ceptor CXCR4 is required for outgrowth of colon carcinoma micrometastases. and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121: 335–348. Cancer Res. 63: 3833–3839. 46. Sawada, S., K. Gowrishankar, R. Kitamura, M. Suzuki, G. Suzuki, S. Tahara, and 23. Moran, C. A. 2006. Pulmonary adenocarcinoma: the expanding spectrum of his- ϩ tologic variants. Arch. Pathol. Lab. Med. 130: 958–962. A. Koito. 1998. Disturbed CD4 T cell homeostasis and in vitro HIV-1 suscep- 24. Wald, O., O. Pappo, R. Safadi, M. Dagan-Berger, K. Beider, H. Wald, tibility in transgenic mice expressing T cell line-tropic HIV-1 receptors. J. Exp. S. Franitza, I. Weiss, S. Avniel, P. Boaz, et al. 2004. Involvement of the Med. 187: 1439–1449. CXCL12/CXCR4 pathway in the advanced liver disease that is associated with 47. White, E. S., K. R. Flaherty, S. Carskadon, A. Brant, M. D. Iannettoni, J. Yee, hepatitis C virus or hepatitis B virus. Eur. J. Immunol. 34: 1164–1174. M. B. Orringer, and D. A. Arenberg. 2003. Macrophage migration inhibitory 25. Avniel, S., Z. Arik, A. Maly, A. Sagie, H. B. Basst, M. D. Yahana, I. D. Weiss, factor and CXC chemokine expression in non-small cell lung cancer: role in B. Pal, O. Wald, D. Ad-El, et al. 2006. Involvement of the CXCL12/CXCR4 pathway angiogenesis and prognosis. Clin. Cancer Res. 9: 853–860. in the recovery of skin following burns. J. Invest. Dermatol. 126: 468–476. 48. Woo, E. Y., C. S. Chu, T. J. Goletz, K. Schlienger, H. Yeh, G. Coukos, ϩ ϩ 26. Sancho, D., M. Gomez, F. Viedma, E. Esplugues, M. Gordon-Alonso, S. C. Rubin, L. R. Kaiser, and C. H. June. 2001. Regulatory CD4 CD25 T cells M. A. Garcia-Lopez, H. de la Fuente, A. C. Martinez, P. Lauzurica, and in tumors from patients with early-stage non-small cell lung cancer and late-stage F. Sanchez-Madrid. 2003. CD69 downregulates autoimmune reactivity through ovarian cancer. Cancer Res. 61: 4766–4772. active transforming growth factor-␤ production in collagen-induced arthritis. 49. Woo, E. Y., H. Yeh, C. S. Chu, K. Schlienger, R. G. Carroll, J. L. Riley, L. R. Kaiser, J. Clin. Invest. 112: 872–882. and C. H. June. 2002. Cutting edge: regulatory T cells from lung cancer patients 27. Esplugues, E., D. Sancho, J. Vega-Ramos, C. Martinez, U. Syrbe, A. Hamann, directly inhibit autologous T cell proliferation. J. Immunol. 168: 4272–4276. P. Engel, F. Sanchez-Madrid, and P. Lauzurica. 2003. Enhanced antitumor im- 50. Franitza, S., O. Kollet, A. Brill, G. G. Vaday, I. Petit, T. Lapidot, R. Alon, and munity in mice deficient in CD69. J. Exp. Med. 197: 1093–1106. O. Lider. 2002. TGF-␤1 enhances SDF-1␣-induced chemotaxis and homing of 28. Zheng, S. G., J. H. Wang, J. D. Gray, H. Soucier, and D. A. Horwitz. 2004. naive T cells by up-regulating CXCR4 expression and downstream cytoskeletal Natural and induced CD4ϩCD25ϩ cells educate CD4ϩCD25Ϫ cells to develop effector molecules. Eur. J. Immunol. 32: 193–202.