| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Pediatrics,
Neurology, and
Radiology, and
Immunology Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Another family includes the UL16 binding proteins, which originally were identified by their binding to human cytomegalovirus protein UL16 (7). MIC homologs have not been identified in mice. The retinoic acid early inducible proteins and H60, however, serve a similar function in this species (8, 9, 10). It is now recognized that MIC, UL16 binding protein, retinoic acid early inducible protein, and H60 are all ligands for the activating NKG2D receptor, which in humans is expressed on most NK cells, NKT cells, and CD8 T cells (11, 12, 13).
Ectopic expression of NKG2D ligands induces rejection of tumors by NK cells and CTLs in syngeneic mice (14, 15). MIC Ags are broadly expressed on epithelial tumors, and their expression is associated with an increased infiltration of NKG2D+ lymphocytes (6). Engagement of NKG2D by MIC enhances cytolytic and cytokine responses of CD8 T cells (16). Despite this, many MIC-expressing tumors rapidly progress, implying that NKG2D-MIC interactions are impaired. It has recently been demonstrated that tumor-derived soluble MIC (sMIC) down-regulates expression of NKG2D on CD8 T cells and suppresses T cell activation (17). However, the effects of down-regulation of NKG2D-mediated activation signals on NK cell effector function in vivo remains to be explored.
Another group of activating NK cell receptors, the natural cytotoxicity receptors (NCRs) NKp46, NKp30, and NKp44, recognize tumor cells in an MHC-independent manner. The ligands for the NCRs are not elucidated, although they are thought to recognize certain tumors of the neuro-epithelial origin (18). Putative viral ligands for the NCR have also been reported (19). The majority of peripheral blood NK cells constitutively express NKp46 and NKp30 but not NKp44, which is up-regulated upon activation with IL-2 (20).
Migratory and homing characteristics of the NK cells will depend upon the expression of a variety of specific chemokine receptors (21, 22, 23). The CC chemokine receptor CCR7, which is important for T cell homing into secondary lymphoid organs, is not expressed on resting NK cells, but is up-regulated in activated NK cells. CCR7 is a shared receptor for CCL19 and CCL21 and participates in the recruitment and proliferation of activated NK cells (24). Another CC chemokine receptor, CCR5, is not expressed in either unactivated or activated NK cells (21, 22, 23). NK cells also express the CXC receptors, i.e., CXCR1, which binds to IL-8, and CXCR2, which binds to epithelial-neutrophil-activating peptide (ENA-78) (25). High levels of ENA-78 in fresh lung cancer specimens have correlated with tumor vascularity and spontaneous metastasis (26). IL-8 is also frequently produced by highly metastatic tumors (27, 28) and potentiates release of matrix metalloproteinases (MMPs) from the tumor cell (29), which in turn can enhance IL-8 activity by proteolytic cleavage (30).
MMPs are synthesized and secreted by cancer cells and by adjacent stromal cells and play important roles in degradation of extracellular matrix proteins (31, 32, 33). It was recently demonstrated that MMPs also release MICA/B from tumor cells (34). MMP-2 and MMP-9, the two secreted MMPs, are expressed at high levels in metastatic tumors (31, 32, 33). Taken together, these findings would suggest that tumor cell evasion from NK cell immunity during tumor development could be mediated by the combined effects of the release of MIC Ags and CXC ligands from tumor cells.
Here we test the hypothesis that some human epithelial tumors evade NK cell-mediated immune responses by down-modulating functionally important NK receptors by release of soluble decoy ligands. We demonstrate, for the first time in NK cells, that the soluble NKG2D ligands present in the sera of patients with colorectal cancer down-modulate both the activating and chemokine receptor groups, resulting in inability of NK cells to target the tumor and inactivation of NK cell effector functions in vitro and in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FACS-purified NK cells were isolated from PBMCs of three normal donors and three patients with colorectal cancer. For patient number 1 (localized primary tumor), samples were obtained at diagnosis and at 29, 32, and 38 mo thereafter. For patient number 2 (metastatic lung nodules), two samples were obtained 5 mo apart. For patient number 3 (metastatic disease in the peritoneum and liver), two samples were obtained 4 mo apart. Written informed consent was obtained from all patients and normal donors per the guidelines of the institutional review board. For colorectal tumor cells, fresh cell suspensions of surgically removed tumors from patients 1 and 2 were prepared by standard techniques. The EBV-transformed B-lymphoblastoid cell line (BLCL) was generated from patients 1 and 2 (i.e., autoBLCL) by standard techniques.
Antibodies
Primary. Murine anti-human CD56 and CD3 (BD Biosciences, Mountain View, CA), goat polyclonal anti-human NKG2D and MIC (Santa Cruz Biotechnology, Santa Cruz, CA), and mouse anti-human CXCR1, CCR7, and CCR5 (R&D Systems, Minneapolis, MN) Abs were purchased. F(ab')2 fragments of anti-NKG2D Ab were generated after pepsin digestion of goat polyclonal NKG2D Ab per the standard procedure. The purity of the F(ab')2 fragment was determined after ascertaining the correct band size on gel electrophoresis. Anti-NKp46 and anti-NKp44 Abs were a kind gift from Dr. Marco Colonna (Washington University School of Medicine, St. Louis, MO).
Secondary. Affinity-purified second Abs and species-absorbed conjugates for dual-labeling were purchased from Chemicon International (Temecula, CA).
In vitro culture conditions
Condition A: PBMC NK cells were directly used for assays. Condition B: PBMC NK cells were cocultured with irradiated autologous PBMCs and EBV-BLCL in pooled human serum (from blood group AB individuals) and activated from day 7–21 with IL-2 (provided by the National Cancer Institute/Biological Response Modifiers Program, Frederick, MD) as previously described (35). Condition C: PBMC NK cells were expanded in culture condition B until day 14, when human AB serum was replaced with autologous patient serum. For the three normal donors, NK cells were propagated from 0–21 days in serum obtained from patient number 1. Condition D: PBMC NK cells were expanded in culture condition B until day 14, when human AB serum was replaced with autologous patient serum pretreated for 1 h with anti-MIC Ab at 1/10 dilution. Condition E: PBMC NK cells were expanded in culture condition B until day 14, when human AB serum was replaced with autologous patient serum pretreated with anti-IgG isotype control Ab at 1/10 dilution. Condition F: PBMC NK cells were cultured from day 0–21 in autologous patient serum and activated with IL-2 from day 7–21. NK cells cultured in each of these conditions were fed with fresh medium, IL-2, and the designated serum every 2–3 days. For conditions B–F, NK cells were analyzed on day 21.
Recombinant human ENA-78 and IL-8 used for studying the temporal dynamics of CXCR1 internalization in NK cells were purchased from R&D Systems. For 2 and 24 h at 37°C, we exposed CXCR1+ NK cells to two concentrations of rIL-8, i.e., low dose (1300 pg/ml; highest concentration obtained in the patient serum) and high dose (1100 ng/ml; concentration that is known to internalize CXCR1 in neutrophils). Similarly, 1000 pg/ml and 1100 ng/ml concentrations of rENA-78 were used. NK cells were analyzed by fluorescent microscopy.
Flow cytometry
Two-color FACS analysis was performed on NK cells labeled with anti-CD56 and either anti-NKG2D or anti-CXCR1 Abs, and the tumor cell suspensions were labeled with either anti-MIC or anti-CD56 Ab using a FACScan cytometer (BD Biosciences).
Cytotoxicity assay
Cell killing assays were performed using NK cells as effectors and 51Cr-labeled autologous EBV-BLCL, autologous colorectal cancer cells, and HLA class I-negative EBV-BLCL 721.221 as targets in the presence or absence of F(ab')2 fragment of NKG2D Ab. Assays were performed in triplicates for 4h at an effector : target ratio of 10:1 and the percent specific lysis calculated as per the standard formula.
ELISA
ELISA for sMICA/B. Primary polyclonal goat anti-human MICA/B Ab (Santa Cruz Biotechnology) was added in 1/100 dilution to the plate precoated with the test serum or PBS and incubated for 1 h at room temperature. Secondary HRP-conjugated rabbit anti-goat IgG (Zymed Laboratories, San Francisco, CA) was added at 1/1000 dilution and incubated at room temperature. After 1 h, substrate reagent (R&D Systems) was added. The reactions were read at adsorption wavelengths of 450 nm and 570 nm using a microplate reader (SAFIRE, Tecan, Austria). Assays were performed in quadruplets and from at least two separate samples.
ELISA for sMMPs, ENA-78, and IL-8. Commercially available ELISA kits (R&D Systems) were used to quantitate serum levels of MMP-2, MMP-9, ENA-78, and IL-8 per the manufacturer’s protocols. ELISA samples were read at a wavelength of 450 nm with a wavelength correction of either 570 or 630 nm, and OD values were used to construct a standard curve. Assays were performed in duplicate and from at least two to four separate samples.
SCID mouse experiment
Equal numbers of EBV-BLCL and colon adenocarcinoma cells (i.e., 5 x 106) derived from patient number 1 were inoculated s.c. in the left and right thighs of the SCID mice (Taconic Farms, Germantown, NY), based on the previously described model (36). Animals were divided into three groups. Group 1 (n = 7) was treated with autologous NKG2D+ NK cells expanded in culture condition B. Group 2 (n = 9) was treated with autologous NKG2D- NK cells expanded in culture condition F. Group 3 (n = 12) received no NK cells. Treatment of the animals was begun only when the tumor size reached 
0.25 cm3 in each of the three groups. NK cells were injected i.v. on days 6, 21, and 43 after tumor inoculation in a total dose of 8 x 106/mouse/course. All animals were treated with 2000 IU of IL-2/mouse i.p. three times a week from days 6–43. Therapeutic effect of NK cells was assessed by measuring tumor volume using formula V = 
x Davg3/6, where Davg is the average diameter (cm) and V is the tumor volume (cm3). The p values were calculated using the Wilcoxon nonparametric statistical analysis. In separate parallel experiments, three animals in each of the above groups were sacrificed on day 5 after the NK cell infusion to determine the homing characteristics of the injected NK cells. Fresh cell suspensions from adenocarcinoma and BLCL as well as liver, spleen, peripheral blood, and bone marrow were evaluated for the presence of NK cells.
Immunohistochemical and immunofluorescent analysis
Tumor samples were fixed in 10% formalin and paraffin embedded by standard techniques. After deparaffinization, sections of each sample were incubated overnight with mouse anti-human CD56 Ab (Zymed Laboratories) at 4°C followed sequentially by incubation for 1 h at room temperature with the biotinylated secondary Ab and peroxidase-conjugated streptavidin. The peroxidase staining was developed using diaminobenzidine (Sigma-Aldrich, St. Louis, MO). Counterstaining was performed using Harris modified hematoxylin (Fisher, Pittsburgh, PA). Monoclonal horse anti-mouse Ab (Vector Laboratories, Burlingame, CA) was used as secondary Ab. All sections were processed with a Mouse-on-Mouse Iso-IHC Kit (Innogenex, San Ramon, CA).
To assess in situ tumor cell-NK cell conjugates by fluorescent microscopy, the deparaffinized samples prepared as above were fixed with 3% paraformaldehyde and incubated for 1 h with primary Ab suspended in 0.01% saponin/0.25% gelatin/0.02% NaN3 (in PBS). After multiple washes, secondary Ab was applied and incubated for 1 h. Slides containing the labeled tissue sections were mounted with a ProLong antifade kit (Molecular Probes, Eugene, OR) and analyzed using a Zeiss Axiovert 200M inverted microscope (New York, NY) with SlideBook software (Intelligent Imaging Innovations, Denver, CO), as previously described (35).
Intracellular and surface staining of NKG2D, MICA/B, and CXCR1 was performed as described previously (35).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NK cells obtained from PBMCs of three patients with colorectal cancer were tested by flow cytometry for expression of NKG2D. Only 3–7% of the NK cells were NKG2D+, whereas 60–67% of NK cells from normal donors expressed NKG2D (Table IA). In the patients, decreased numbers of NKG2D+ NK cells were observed in multiple samples obtained through the course of the disease.
|
|
70%) displayed intracellular MIC; few NK cells (5–10%) displayed surface expression of MIC (Fig. 2b). The presence of MIC on the surface or within the cytoplasm of the NK cell suggests the presence of tumor-derived circulating sMIC in the patients with colorectal cancer.
|
|
The effect of modulation of NKG2D on NK cell-mediated cytotoxicity in vitro was analyzed by testing the tumoricidal activity of NK cells from patients 1 and 2 (data not shown for patient number 1). NK cytotoxicity against MIC+ autologous tumor cells was not observed with PBMC NK cells and with NK cells cultured in sMIC-containing autologous serum (Fig. 1, A and C), both of which were predominantly NKG2D-. However, NK cells cultured in either normal AB serum (Fig. 1B) or autologous patient serum pretreated with anti-MIC Ab (Fig. 1D) became NKG2D+ and were now cytotoxic to the tumor cells. Furthermore, F(ab')2 fragment of anti-NKG2D Ab blocked antitumor cytotoxicity by NKG2D+ NK cells, whereas 721.221 cells remained susceptible (Fig. 1B, asterisk). NK cells cultured in autologous serum pretreated with IgG isotype control Ab failed to up-regulate NKG2D and failed to lyse carcinoma cells (Fig. 1D, star). Because 721.221 cells were satisfactorily lysed by IL-2-activated patient NK cells and because the activating receptor NKp46 is thought to mediate this killing (20), we determined the expression profile of NKp46 and NKp44 in circulating and IL-2-activated patient NK cells (Table I). Although NKp46 was absent from PBMC NK cells of the patients, this receptor became up-regulated after propagation in IL-2 independently of the serum source, i.e., from healthy donors or patients (Table I, B and C). As expected, NKp44 was not expressed in the circulating PBMC NK cells from healthy donors or patients but became expressed on a large fraction of normal NK cells and NK cells from the patients when cultured in human AB serum for 21 days (Table IB). However, when the normal or patient NK cells were cultured in MIC-containing serum, NKp44 became down-regulated (Table IC). In contrast, large numbers of NK cells continued to express NKp44 when cultured in MIC-containing patient serum coincubated with anti-MIC Abs but not with isotype control Ab (Table I, D and E).
Taken together, these studies demonstrate that sMIC in the serum of the patients with colorectal cancer binds to NKG2D, leading to its internalization. The resultant reduced expression of NKG2D leads to impaired tumoricidal activity against MIC-expressing tumor cells in vitro, which require NKG2D activation for their cytotoxicity.
Expression of NKG2D determines the capacity of NK cells to prevent growth of MIC+ human colorectal cancer in SCID mice
The in vivo effects of NKG2D- and NKG2D+ NK cells on tumor growth were evaluated in SCID mice. Animals xenografted with the tumor from patient number 1 were treated with 1) autologous NKG2D+ NK cells, 2) NKG2D- NK cells, or 3) IL-2 alone (Fig. 4a, left panel). NKG2D+ NK cell-treated animals demonstrated significantly smaller tumor size at each measurement done weekly over the period of observation, compared with either NKG2D- NK cell-treated or IL-2-treated animals. However, there was no difference in the growth of MIC- EBV-BLCL tumor derived from the same patient, which was concurrently implanted in the opposite flank of the animals (Fig. 4a, right panel).
|
Tissue sections of the xenografted carcinoma and EBV-BLCL tumors were also examined by immunohistochemisty and in situ digital immunofluorescence microscopy. Adenocarcinomas treated with NKG2D+ NK cells demonstrated infiltration of infused NK cells within the tumor substance, resulting in disruption of the tumor architecture and areas of necrosis (Fig. 4c, top left and middle panels), whereas this was not seen in mice treated with NKG2D- NK cells or IL-2 alone (data not shown). In addition, NK cell infiltration was not observed in the EBV-BLCL tumors treated with either NKG2D+ or NKG2D- NK cells (Fig. 4c, top right panel).
Similarly, immunofluorescence analysis of carcinoma xenografts demonstrated NK cell infiltration only in animals treated with NKG2D+ NK cells (Fig. 4c, bottom left panel). Furthermore, multiple NK cell-carcinoma cell conjugates demonstrated polarization of NKG2D to the cell-cell contact area, indicating formation of an activating NK cell immune synapse in situ (Fig. 4c, bottom right panel).
These in vivo experiments demonstrate significant growth retardation of implanted colorectal tumors when treated with autologous NKG2D+ NK cells, whereas mice treated with NKG2D- NK cells or IL-2 alone had tumor progression. Strikingly, whereas NKG2D+ NK cells infiltrated carcinoma xenografts, NKG2D- NK cells failed to do so.
NK cells from patients with MIC-expressing colorectal cancer lack expression of the chemokine receptors CXCR1 and CCR7
Because the infused NKG2D- NK cells failed to localize to the MIC+ tumors, we analyzed cell surface expression of chemokine receptors important for cell migration and homing to inflammatory sites such as CXCR1, CCR5, and CCR7 (21, 22, 23, 24). As expected, PBMC NK cells from the three patients and the normal donors had no significant expression of CCR7 (Table IA) or CCR5. Although most normal NK cells expressed CXCR1, the proportion of circulating NK cells from the patients expressing this receptor was very low (Table IA). Cell membrane expression of CXCR1 was below detectable levels by flow cytometry in the majority of NK cells. Immunofluorescence analysis of NK cells from the patients demonstrated increased CXCR1 in the cytoplasm (Fig. 2c). In contrast, NK cells from normal individuals expressed CXCR1 on the cell membrane. When the patient NK cells were cultured for 21 days with IL-2 and normal human AB serum, the number of CXCR1+ NK cells dramatically increased (Table IB). Whereas culture of NK cells in the identical condition for 14 days induced membrane expression of CXCR1, subsequent transfer to autologous patient serum led to dramatic reduction in the number of CXCR1+ NK cells (Table IC). CXCR1 was also down-modulated on normal NK cells cultured in serum from patient number 1 (containing high sMIC and ENA-78 but low IL-8 levels), contrasting with sustained high expression observed upon culturing in normal serum (Table IC). These findings parallel our studies of the modulation of NKG2D expression. In addition, however, the down-modulation of CXCR1 did not occur when NK cells from the patients were cultured in autologous serum coincubated with anti-MIC Ab (Table I, D and E).
The CC chemokine receptor CCR7, which is expressed on activated NK cells, became expressed on a large fraction of normal NK cells and NK cells from the patients after culturing with IL-2 in human AB serum for 21 days (Table IB). In contrast, when the normal or patient NK cells were cultured in patient serum, this activation marker was down-regulated (Table IC). In contrast, the high level of expression of CCR7 was sustained when these NK cells were cultured in patient serum coincubated with anti-MIC Abs (Table I, D and E). Immunofluorescence analysis of IL-2-activated CCR7- patient NK cells that were cultured in patient serum revealed strictly a cytoplasmic localization of CCR7 with no signals detected in the membrane, which is in contrast with normal IL-2-activated NK cells in which CCR7 is predominantly membrane localized with only a small fraction present in the cytoplasm close to the microtubule organizing center (data not shown).
To further study the mechanism(s) of CXCR1 modulation in NK cells, we determined whether increased serum levels of the CXC ligands were responsible for receptor-mediated down-modulation. Therefore, serum levels of IL-8 and ENA-78 were quantitatively analyzed by ELISA. The levels of IL-8 were significantly elevated in patients 2 and 3, whereas ENA-78 was increased in each of the three patients (Fig. 3b). Because MMP-9 is known to increase IL-8 activity in the tumor microenvironment (30) and both MMP-2 and MMP-9 are secreted by colon adenocarcinomas (31, 32), we tested the serum levels of these enzymes and found both markedly elevated in each of the patients studied (Fig. 3b, top panels).
Because incubating NK cells in serum (patient number 1) not containing high IL-8 resulted in internalization of CXCR1 and pretreatment of this serum with anti-MIC resulted in up-regulation of CXCR1, we designed experiments to address whether it is the CXC ligands or NKG2D ligands that are involved in internalization of CXCR1. Therefore, CXC ligand-mediated modulation of CXCR1 on NK cells was evaluated in in vitro experiments. Because IL-8-induced receptor internalization has been demonstrated for CXCR1 in the neutrophils, we exposed CXCR1+ NK cells to low and high concentrations of rIL-8 and rENA-78 for 2 and 24 h as described in Materials and Methods. FACS and immunofluorescence analysis of NK cells incubated with either ligand at both doses and time points failed to demonstrate a decrease in the number of CXCR1+ NK cells or decreased surface expression and/or internalization of CXCR1 (data not shown).
Having demonstrated no role for IL-8 and ENA-78 in receptor internalization and because CXCR1 expression was influenced by the absence or presence of sMIC in the culture medium, we then determined the in vitro effect of sMIC-containing serum on the temporal relationship between NKG2D and CXCR1 expression. CXCR1 and NKG2D-expressing IL-2-activated NK cells obtained from a normal donor were incubated for 4, 24, 65, and 140 h in medium containing serum from patient number 3 that contained a sevenfold higher concentration of sMIC than in sera from normal controls. In these experiments, as opposed to the experiment detailed in Table I, the medium containing the serum was not changed over the 6 days of culture. The NK cells were then analyzed by immunofluorescent microscopy (Fig. 5). At 4 h, the majority of NK cells demonstrated internalization of NKG2D but not of CXCR1. At 24 h, half of the NK cells showed almost complete absence of NKG2D, whereas the other half showed some degree of internalization of NKG2D. Interestingly, CXCR1 was seen internalized only in those NK cells in which NKG2D was completely absent, whereas normal surface expression of CXCR1 was observed in the NK cells that contained NKG2D either on the surface or in the cytoplasm. At 65 h, 20% of the NK cells showed complete absence of NKG2D and internalization of CXCR1, whereas the rest of the NK cells showed internalization of NKG2D only. By 6 days, the majority of NK cells regained normal surface expression of NKG2D and CXCR1 similar to that of untreated normal NK cells, thereby reflecting depletion of sMIC from the medium over the course of culture (Fig. 5).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Functional deficiency of T cells, i.e., cytotoxicity and IFN-
production in vitro, has been shown to correlate with expression of NKG2D (17). Our results are the first to demonstrate functional impairment of the adoptively transferred NKG2D- NK cells in vivo.
The ability of NK cells to selectively traffic to tumor sites will depend on the chemokine receptor expression profile and chemokine responsiveness, both of which are different from those of T cells (22). Our studies have demonstrated that NKG2D-CXCR1- and CCR7- infused NK cells were not detected in the carcinoma xenografts of the treated mice, whereas NKG2D+CXCR1+ and CCR7+ NK cells were found infiltrating carcinoma xenografts. Whether this phenomenon is due to failure to traffic (chemokine receptor/chemokine defect) to the tumor sites or failure to remain adhered to the tumor cells (adhesion defect) after successfully trafficking cannot be ascertained from our experiments, although we did observe elevated levels of soluble ICAM-1 in the sera of patients with colorectal cancer (our unpublished observations), as have others (37), suggesting that multiple mechanisms must play a role in failure to home to tumor sites.
Release of tumor-derived soluble ligands for important receptors on the immune cells is an effective mechanism of evading tumor immunity, as is depleting the tumor cell surface of the molecular targets for NK cells or CTLs. MMP-mediated proteolytic shedding of MIC molecules is at least one of the mechanisms involved in the release of sMIC in the circulation and consequent reduced expression of MIC on the surface of the tumor cell (34). Although we did not attempt to correlate surface density of MIC molecules with serum levels of sMIC, one recent study failed to demonstrate such a relationship (38). We hypothesize that production of MMPs in the tumor microenvironment leads to proteolytic cleavage and release of NKG2DLs, particularly MIC molecules. Alternatively, MIC molecules can be secreted into the tumor microenvironment as alternatively spliced variants (39). In either case, soluble ligand(s) would bind to NKG2D, inducing receptor internalization and abolishing NK cell effector functions. Secretion of tumor-derived sMIC is not confined to solid epithelial tumors but is also seen with leukemia cells in vivo (38).
Internalization of NK receptor expression in vivo and in vitro is not limited to NKG2D but is also observed for the chemokine receptors CXCR1 and CCR7. The immunofluorescence analysis of circulating PBMC NK cells clearly demonstrated internalization of CXCR1. Although serum levels of the CXC glutamic acid-leucine-arginine+/- chemokines IL-8, a strong agonist for CXCR1, and ENA-78, a weak agonist for CXCR1 (P.M. Murphy, unpublished observations), were elevated in the patients, our studies indicate that down-modulation of CXCR1 is probably not mediated by these soluble CXC ligands, at least at the time points studied. These findings are in contrast with neutrophils, in which dose-dependent IL-8-mediated receptor internalization is observed (40). However, we cannot exclude the possibility that incubation with these and other (i.e., granulocyte chemotactic protein 2) (41) chemokine ligands at additional time points, earlier or later, may result in desensitization and internalization of CXCR1.
The mechanism(s) responsible for up-regulation of CXCR1 on NK cells incubated in patient sera pretreated with anti-MIC is less clear. Studies on the temporal kinetics of sMIC-induced phosphorylation-dephosphorylation of the chemokine receptors are likely to throw additional light on the mechanism for receptor internalization (40, 41, 42). Our in vitro experiments do demonstrate, however, a temporal relationship between expression of NKG2D and CXCR1, and they also show modulation of receptor expression of both NKG2D and CXCR1 by sMIC. Upon incubating NK cells with sMIC-containing serum, we observed the first indications of internalization of NKG2D at 4 h and maximum internalization by 24 h, with over half of the analyzed NK cells showing total absence of NKG2D. These results are in agreement with the observations made in CD8+ T cells incubated with purified recombinant sMICA (17). In contrast, internalization of CXCR1 was not observed before 24 h of incubation with sMIC-containing serum. It is of note that CXCR1 was internalized only in those NK cells with complete internalization and no surface expression of NKG2D. Partial internalization with residual surface expression of NKG2D was not enough to modulate expression of CXCR1. These results and results from the temporal analysis of NKG2D and CXCR1 internalization would suggest that sMIC-mediated down-modulation of the triggering receptor NKG2D is rapid and precedes down-modulation of chemokine receptors. Our results also imply that reduction in surface density of NKG2D must reach a critical threshold before secondary effects on other receptors can be observed.
Furthermore, we observed down-modulation in the IL-2 activated patient NK cells of other activation markers, such as the CC chemokine receptor CCR7 and the NCR NKp44, which are normally up-regulated in IL-activated NK cells. A causal relationship between soluble ligand-mediated down-modulation of NKG2D and down-modulation of other receptors has not been established, but may reflect NKG2D-mediated general inactivation of the NK cells. Similar findings of activating receptor down-modulation in NK cells from patients with acute leukemia have been described (43).
Alternatively, NKG2D-DNAX-activating protein 10 (DAP10)-mediated activation pathways in humans may be involved in modulation of CXCR1 and CCR7, and extinguishing the NKG2D-DAP10 signals will not only prevent NKG2D-mediated cytotoxicity but will also perturb chemokine receptor expression. The observation that the triggering receptor expressed on myeloid cell-2/DAP12-mediated activation pathway regulates the expression of CC chemokine receptor CCR7 in murine dendritic cells is another example of cross-talk between activation and chemokine receptors (44).
It is important to emphasize that the observed down-modulation of NKp44 and the chemokine receptors are events that are perhaps secondary to the down-modulation of NKG2D and are likely not directly caused by physical association with sMIC present in the sera of the patients.
Our in vitro studies demonstrate re-expression of NKG2D and recovery of NK cytotoxicity in vitro after culture of NK cells in normal serum. Our observation that selective inhibition of NK cytotoxicity against MIC+ tumor cells, but not 721.221 cells, occurs upon blocking NKG2D receptor with anti-NKG2D-F(ab')2 clearly indicates that functional NKG2D is necessary for tumor cell killing in vitro in these autologous combinations. These findings are in agreement with studies in mice bearing tumors transduced to express NKG2D ligands (NKG2DL) (14, 15). In these models, NK cells were able to reject tumors despite their expression of self-MHC class I.
It is of particular interest that NKG2D-expressing NK cells, in our studies, were able to mediate effector responses against tumors that constitutively express NKG2DL.
Our study is the first in demonstrating in vivo polarization of NKG2D to cell-cell contact area and formation of productive immune synapse in the in situ NK cell-tumor cell conjugates.
Recently, reversible down-modulation of NKG2D upon exposure to NKG2DL was seen in NK cells from nonobese diabetic mice, resulting in impairment of NK cytotoxicity and cytokine production (45). These findings support the notion that the NKG2D-NKG2DL receptor-ligand pair is a powerful regulator of NK cell effector functions. The prolonged changes in NKG2D expression and slow reappearance are in agreement with observations made in a recent study of TGF
1-induced modulation of NKG2D expression in NK cells. Here the decreased surface expression of NKG2D correlated with the decrease in NKG2D transcript expression in some but not in all NK clones or in polyclonal NK cells (46). This study also demonstrated regulation of triggering receptor expression on NK cells by cytokine TGF-, which is secreted by a variety of malignant tumors and imparts an aggressive phenotype (47, 48, 49).
In the experiments presented in Table I, the medium containing patient serum was replaced every 2–3 days, thus maintaining high concentrations of sMIC all through 21 days of culture. That NKG2D can be re-expressed by NK cells is illustrated by the effect of a single pulse of medium containing patient serum on normal NK cells described in Fig. 5. In this experiment, the NK cells gradually regained surface expression of NKG2D between 65 and 140 h of culture. Whether re-expression of NKG2D is a result of receptor recycling or de novo protein synthesis in the NK cells or their progeny cannot be ascertained from our experiments. However, a recent study in CD8 T cells demonstrated that MIC-mediated internalized NKG2D was not recycled but was degraded (17).
The described tumor escape mechanisms involving production and release of soluble decoy ligands for lymphocyte receptors are very similar to processes used by viruses in their evasion of immune responses (50, 51, 52). It should also be emphasized that the adaptive immune response to MIC-expressing tumors such as breast, colon, prostate, and lung cancers as well as leukemias (38) can be evaded by mechanisms similar to escape from NK immunity because CD8+ T cells also use NKG2D as an activating receptor (35, 53, 54). Our ongoing studies have shown that in other epithelial tumors such as ovarian cancer, the circulating PBMC-derived NK cells were also significantly deficient of NKG2D (data not shown).
Although it is intriguing how tumors have developed a mechanism(s) to evade the immune system by receptor perturbation mediated by soluble decoy ligands, the timing and the mechanism of shedding of the surface MIC by tumor cells during tumor genesis will be interesting to explore. It would be especially interesting to determine whether this requires a large tumor cell mass or reflects specific changes in the tumor cells. It is possible that early tumor cells do not shed this ligand, making them an easy target for NK and T cell cytolytic function, whereas secondary mutations in the tumor cells are responsible for this effect. Our studies stress the need to carefully select lymphocytes, based on their phenotypic characteristics, for successful implementation of adoptive cell-based therapies as a means of treating cancers. Alternatively, manipulation of enzymes responsible for generation of soluble decoy ligands may offer attractive future approaches.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Yatin M. Vyas, Box 376, Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. E-mail address: vyasy{at}mskcc.org
3 Abbreviations used in this paper: MIC, MHC class I chain-related molecule; sMIC, soluble MIC; NCR, natural cytotoxicity receptor; ENA-78, epithelial-neutrophil-activating peptide 78; MMP, matrix metalloproteinase; BLCL, B-lymphoblastoid cell line; DAP10, DNAX-activating protein 10.
Received for publication August 8, 2003. Accepted for publication October 6, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cells. Science 294:605.
T cells by NKG2D via engagement by MIC induced on virus-infected cells. Nat. Immunol. 2:255.[Medline]
and leaves RANTES and MCP-2 intact. Blood 96:2673.1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells. Proc. Natl. Acad. Sci. USA 100:4120. signaling in tumor suppression and cancer progression. Nat. Genet. 29:117.[Medline]
(TGF-1) gene by human T lymphotropic virus type 1 tax: a potential mechanism for the increased production of TGF- 1 in adult T cell leukemia. J. Exp. Med. 172:121.1 increases survival of human melanoma through stroma remodeling. Cancer Res. 61:8306. T cells. J. Immunol. 169:1236.Related articles in The JI:
This article has been cited by other articles:
|
|
 |
|
  F. Romagne, P. Andre, P. Spee, S. Zahn, N. Anfossi, L. Gauthier, M. Capanni, L. Ruggeri, D. M. Benson Jr, B. W. Blaser, et al. Preclinical characterization of 1-7F9, a novel human anti-KIR receptor therapeutic antibody that augments natural killer-mediated killing of tumor cells Blood, September 24, 2009; 114(13): 2667 - 2677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-N. Peraldi, J. Berrou, N. Dulphy, A. Seidowsky, P. Haas, N. Boissel, F. Metivier, C. Randoux, N. Kossari, A. Guerin, et al. Oxidative Stress Mediates a Reduced Expression of the Activating Receptor NKG2D in NK Cells from End-Stage Renal Disease Patients J. Immunol., February 1, 2009; 182(3): 1696 - 1705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Wu, C. L. Atteridge, X. Wang, T. Seya, and S. R. Plymate Obstructing Shedding of the Immunostimulatory MHC Class I Chain-Related Gene B Prevents Tumor Formation Clin. Cancer Res., January 15, 2009; 15(2): 632 - 640. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Boutet, S. Aguera-Gonzalez, S. Atkinson, C. J. Pennington, D. R. Edwards, G. Murphy, H. T. Reyburn, and M. Vales-Gomez Cutting Edge: The Metalloproteinase ADAM17/TNF-{alpha}-Converting Enzyme Regulates Proteolytic Shedding of the MHC Class I-Related Chain B Protein J. Immunol., January 1, 2009; 182(1): 49 - 53. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Apps, L. Gardner, J. Traherne, V. Male, and A. Moffett Natural-killer cell ligands at the maternal-fetal interface: UL-16 binding proteins, MHC class-I chain related molecules, HLA-F and CD48 Hum. Reprod., November 1, 2008; 23(11): 2535 - 2548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Waldhauer, D. Goehlsdorf, F. Gieseke, T. Weinschenk, M. Wittenbrink, A. Ludwig, S. Stevanovic, H.-G. Rammensee, and A. Steinle Tumor-Associated MICA Is Shed by ADAM Proteases Cancer Res., August 1, 2008; 68(15): 6368 - 6376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Caligiuri Human natural killer cells Blood, August 1, 2008; 112(3): 461 - 469. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Le Maux Chansac, D. Misse, C. Richon, I. Vergnon, M. Kubin, J.-C. Soria, A. Moretta, S. Chouaib, and F. Mami-Chouaib Potentiation of NK cell-mediated cytotoxicity in human lung adenocarcinoma: role of NKG2D-dependent pathway Int. Immunol., July 1, 2008; 20(7): 801 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Siemens, N. Hu, A. K. Sheikhi, E. Chung, L. J. Frederiksen, H. Pross, and C. H. Graham Hypoxia Increases Tumor Cell Shedding of MHC Class I Chain-Related Molecule: Role of Nitric Oxide Cancer Res., June 15, 2008; 68(12): 4746 - 4753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-J. Kim, M.-K. Han, and H. E. Broxmeyer 4-1BB regulates NKG2D costimulation in human cord blood CD8+ T cells Blood, February 1, 2008; 111(3): 1378 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Giurisato, M. Cella, T. Takai, T. Kurosaki, Y. Feng, G. D. Longmore, M. Colonna, and A. S. Shaw Phosphatidylinositol 3-Kinase Activation Is Required To Form the NKG2D Immunological Synapse Mol. Cell. Biol., December 15, 2007; 27(24): 8583 - 8599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Maccalli, D. Nonaka, A. Piris, D. Pende, L. Rivoltini, C. Castelli, and G. Parmiani NKG2D-Mediated Antitumor Activity by Tumor-Infiltrating Lymphocytes and Antigen-Specific T-Cell Clones Isolated from Melanoma Patients Clin. Cancer Res., December 15, 2007; 13(24): 7459 - 7468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. M. El-Sherbiny, J. L. Meade, T. D. Holmes, D. McGonagle, S. L. Mackie, A. W. Morgan, G. Cook, S. Feyler, S. J. Richards, F. E. Davies, et al. The Requirement for DNAM-1, NKG2D, and NKp46 in the Natural Killer Cell-Mediated Killing of Myeloma Cells Cancer Res., September 15, 2007; 67(18): 8444 - 8449. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Endt, F. E. McCann, C. R. Almeida, D. Urlaub, R. Leung, D. Pende, D. M. Davis, and C. Watzl Inhibitory Receptor Signals Suppress Ligation-Induced Recruitment of NKG2D to GM1-Rich Membrane Domains at the Human NK Cell Immune Synapse J. Immunol., May 1, 2007; 178(9): 5606 - 5611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Osaki, H. Saito, T. Yoshikawa, S. Matsumoto, S. Tatebe, S. Tsujitani, and M. Ikeguchi Decreased NKG2D Expression on CD8+ T Cell Is Involved in Immune Evasion in Patients with Gastric Cancer Clin. Cancer Res., January 15, 2007; 13(2): 382 - 387. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Theriault, H. Adachi, and S. K. Calderwood Role of Scavenger Receptors in the Binding and Internalization of Heat Shock Protein 70 J. Immunol., December 15, 2006; 177(12): 8604 - 8611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Sutherland, B. Rabinovich, N. J. Chalupny, P. Brawand, R. Miller, and D. Cosman ULBPs, human ligands of the NKG2D receptor, stimulate tumor immunity with enhancement by IL-15 Blood, August 15, 2006; 108(4): 1313 - 1319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Roda-Navarro, M. Vales-Gomez, S. E. Chisholm, and H. T. Reyburn Transfer of NKG2D and MICB at the cytotoxic NK cell immune synapse correlates with a reduction in NK cell cytotoxic function PNAS, July 25, 2006; 103(30): 11258 - 11263. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Jinushi, F. S. Hodi, and G. Dranoff Therapy-induced antibodies to MHC class I chain-related protein A antagonize immune suppression and stimulate antitumor cytotoxicity PNAS, June 13, 2006; 103(24): 9190 - 9195. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kim, M. Emi, K. Tanabe, and K. Arihiro Tumor-Driven Evolution of Immunosuppressive Networks during Malignant Progression Cancer Res., June 1, 2006; 66(11): 5527 - 5536. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Boissel, D. Rea, V. Tieng, N. Dulphy, M. Brun, J.-M. Cayuela, P. Rousselot, R. Tamouza, P. Le Bouteiller, F.-X. Mahon, et al. BCR/ABL oncogene directly controls MHC class I chain-related molecule A expression in chronic myelogenous leukemia. J. Immunol., April 15, 2006; 176(8): 5108 - 5116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Mincheva-Nilsson, O. Nagaeva, T. Chen, U. Stendahl, J. Antsiferova, I. Mogren, J. Hernestal, and V. Baranov Placenta-Derived Soluble MHC Class I Chain-Related Molecules Down-Regulate NKG2D Receptor on Peripheral Blood Mononuclear Cells during Human Pregnancy: A Possible Novel Immune Escape Mechanism for Fetal Survival J. Immunol., March 15, 2006; 176(6): 3585 - 3592. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Liu, S. Yu, K. Zinn, J. Wang, L. Zhang, Y. Jia, J. C. Kappes, S. Barnes, R. P. Kimberly, W. E. Grizzle, et al. Murine Mammary Carcinoma Exosomes Promote Tumor Growth by Suppression of NK Cell Function J. Immunol., February 1, 2006; 176(3): 1375 - 1385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Sconocchia, M. Lau, M. Provenzano, K. Rezvani, W. Wongsena, H. Fujiwara, N. Hensel, J. Melenhorst, J. Li, S. Ferrone, et al. The antileukemia effect of HLA-matched NK and NK-T cells in chronic myelogenous leukemia involves NKG2D-target-cell interactions Blood, November 15, 2005; 106(10): 3666 - 3672. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Le Maux Chansac, A. Moretta, I. Vergnon, P. Opolon, Y. Lecluse, D. Grunenwald, M. Kubin, J.-C. Soria, S. Chouaib, and F. Mami-Chouaib NK Cells Infiltrating a MHC Class I-Deficient Lung Adenocarcinoma Display Impaired Cytotoxic Activity toward Autologous Tumor Cells Associated with Altered NK Cell-Triggering Receptors J. Immunol., November 1, 2005; 175(9): 5790 - 5798. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Ghiringhelli, C. Menard, M. Terme, C. Flament, J. Taieb, N. Chaput, P. E. Puig, S. Novault, B. Escudier, E. Vivier, et al. CD4+CD25+ regulatory T cells inhibit natural killer cell functions in a transforming growth factor-{beta}-dependent manner J. Exp. Med., October 17, 2005; 202(8): 1075 - 1085. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Zhang, B. A. Lemoi, and C. L. Sentman Chimeric NK-receptor-bearing T cells mediate antitumor immunotherapy Blood, September 1, 2005; 106(5): 1544 - 1551. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Coudert, J. Zimmer, E. Tomasello, M. Cebecauer, M. Colonna, E. Vivier, and W. Held Altered NKG2D function in NK cells induced by chronic exposure to NKG2D ligand-expressing tumor cells Blood, September 1, 2005; 106(5): 1711 - 1717. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Wiemann, H.-W. Mittrucker, U. Feger, S. A. Welte, W. M. Yokoyama, T. Spies, H.-G. Rammensee, and A. Steinle Systemic NKG2D Down-Regulation Impairs NK and CD8 T Cell Responses In Vivo J. Immunol., July 15, 2005; 175(2): 720 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Smyth, J. Swann, J. M. Kelly, E. Cretney, W. M. Yokoyama, A. Diefenbach, T. J. Sayers, and Y. Hayakawa NKG2D Recognition and Perforin Effector Function Mediate Effective Cytokine Immunotherapy of Cancer J. Exp. Med., November 15, 2004; 200(10): 1325 - 1335. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Roda-Navarro, M. Mittelbrunn, M. Ortega, D. Howie, C. Terhorst, F. Sanchez-Madrid, and E. Fernandez-Ruiz Dynamic Redistribution of the Activating 2B4/SAP Complex at the Cytotoxic NK Cell Immune Synapse J. Immunol., September 15, 2004; 173(6): 3640 - 3646. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
