| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print September 8, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,¶,1
* Departments of Pathology,
Surgery,
Neurobiology, and ¶Medicine, University of Pittsburgh School of Medicine, and University of Pittsburgh Cancer Institute, Pennsylvania; and Department of Oral and Maxillofacial Surgery, University of Mainz, Germany
1 Correspondence: University of Pittsburgh Cancer Institute, Hillman Cancer Center, G.17d, 5117 Centre Avenue, Pittsburgh, PA 15213-1863. E-mail: vujanovicnl{at}msx.upmc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
26% of peripheral blood NK cells but encompass the majority of NK cells in normal and cancerous, solid tissues. We conclude that pre-A-NK cells represent a distinct subset of resting, mature NK cells with the characteristics indicative of their ability to migrate and reside in solid tissues.
Key Words: NK cell subset • pre-A-NK cells • ANK-1 epitope • cell adherence • and tissue NK cells
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(IFN-), IL-4, IL-5, IL-10, IL-13, and transforming growth factor-ß], hematopoietic (e.g., granulocyte macrophage-colony stimulating factor and IL-5), chemotactic [e.g., IL-8, macrophage-inflammatory protein-1
(MIP-1), MIP-1ß, and regulated on activation, normal T expressed and secreted (RANTES)], and antiviral (e.g., MIP-1, MIP-1ß, RANTES, IFN-, and IFN-) cytokines [1
2
3
, 8
9
10
]. Given these multiple functions, NK cells can mediate a variety of important biological responses, including prevention of cancer development, elimination of metastases, protection against infection, initiation of inflammation, regulation of immune responses, regulation of hematopoiesis, and regulation of tissue regeneration [1
2
3
, 7
, 8
, 11
12
13
14
15
].
A remarkable feature of NK cells is their heterogeneity [1
2
3
4
5
6
7
8
, 16
17
18
19
20
21
22
], and it could be possible that the ability of NK cells to mediate multiple different functions is a consequence of their heterogeneity. Two best-defined subpopulations of human NK cells are CD56dimCD16+ and CD56brightCD16– NK cells [21
, 22
]. CD56dimCD16+ NK cells represent 90–95% of peripheral blood NK cells. They express the intermediate affinity IL-2 receptor [IL-2R; ß chain (IL-2Rß)], chemotactic receptors for homing to inflamed tissues, KIRs, and perforin. They mediate NK cytolysis and antibody-dependent, cell-mediated cytotoxicity (ADCC) and secrete low levels of cytokines upon stimulation. In contrast, CD56brightCD16– NK cells constitute only 5–10% of the peripheral blood NK cells. They express the high-affinity IL-2R (IL-2Rß) and CD94/NKG2 but lack killer cell immunoglobulin (Ig)-like receptors. In addition, CD56brightCD16– NK cells express the chemotactic receptor CC chemokine receptor 7 and CD62 ligand for homing to secondary lymphoid tissues, lack perforin, and are noncytotoxic and secrete high levels of IFN- and lymphotoxin (LT)- upon stimulation [21
, 22
].
Based on the selective ability to rapidly (1–5 h) respond to 22 nM IL-2 by a temporal adherence to solid surfaces, we have previously defined two additional, significant subpopulations of NK cells, i.e., activated adherent NK (A-NK) cells and activated nonadherent NK (NA-NK) cells [9
, 11
, 23
24
25
]. Among fresh peripheral blood NK cells, only a small proportion (4–30%) is capable of responding to IL-2 stimulation by adherence. At the time of and immediately after IL-2-induced adherence, A-NK cells notably differ from the rest of IL-2-activated NK (i.e., NA-NK) cells [9
, 11
, 23
24
25
]. A-NK cells produce significant levels of IFN-, IL-1ß, tumor necrosis factor, and LT-, express higher levels of integrins and CD122 (IL-2Rß), and have a higher proliferation rate than NA-NK cells. In addition, A-NK cells express lower levels of CD56 and CD16, mediate lower ADCC, and require longer IL-2 stimulation to develop lymphokine-activated killer (LAK) activity than NA-NK cells (refs. [9
, 23
], unpublished data). The greatest difference between A-NK and NA-NK cells appears to be in their in vitro and in vivo antitumor activities [11
, 24
]. A-NK cells are able to migrate, infiltrate, and kill cancer cells in solid tumor tissues and to eliminate established tumors and metastases. In contrast, NA-NK cells mediate little if any of the anticancer functions. Most of the phenotypic and functional differences of A-NK cells and NA-NK cells are stably maintained in IL-2 cultures [23
]. It is not clear whether A-NK cells represent an activation stage of NK cells in general or represent a distinct subpopulation of NK cells; and the precursors of A-NK (pre-A-NK) cells remain to be defined.
Neural-cell adhesion molecule (N-CAM) is an evolutionarily, highly conserved type I transmembrane glycoprotein of the Ig superfamily, which is predominantly expressed by neural and muscle cells. N-CAM is involved in a variety of important cellular functions, including adhesion, migration, growth, and differentiation [16 , 18 , 26 , 27 ]. Although encoded by a single gene, N-CAM is produced in numerous isoforms as a result of differential splicing of its mRNA and post-translational modifications of its molecule. Different N-CAM isoforms can be selectively and/or differentially expressed and therefore, represent markers of different types of cells and/or their different stages of differentiation [26 ]. In humans, N-CAM is also expressed by NK cells and a small subset of T cells, defined as NK T cells [3 , 16 , 28 29 30 31 ]. Human, fresh, mature NK cells, in contrast to immature NK cells, express CD56 N-CAM [32 ], and mature NK cells differ in the expression levels of CD56 N-CAM, which define CD56dim and CD56bright subpopulations of NK cells [3 , 16 , 21 , 22 , 29 30 31 ]. Therefore, it might be possible that different subpopulations of NK cells or their different stages of maturation are not only defined by different expression levels of CD56 N-CAM but also by differential expression of various N-CAM isoforms.
In the present study, we assessed these possibilities, and attempted to define pre-A-NK cells in fresh human peripheral blood based on selective expression of N-CAM isoforms and/or epitopes. We found that a novel CD56 N-CAM epitope, ANK-1 peptide, is selectively expressed on a small subset of fresh NK cells, which constitutively show increased expression of integrins and IL-2Rß, selective abilities to respond to IL-2 by adherence, and development of A-NK cells with LAK activity and to reside in solid tissues. The remarkable phenotypic and functional similarities of freshly isolated ANK-1+ NK cells and IL-2-induced A-NK cells and the selective ability of ANK-1+ NK cells to differentiate into A-NK cells upon IL-2 stimulation strongly suggest a lineage association in which ANK-1+ NK cells are pre-A-NK cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, clone NCAM16.2), specific for the third extracellular Ig-like domain common to the three main isoforms (120, 140, and 180 kD) of N-CAM, and anti-CD56 (IgG1k, clone MY31, anti-Leu-19) mAb, reactive with the 140-kD CD56 N-CAM isoform expressed on NK cells, were obtained from BD Biosciences PharMingen (San Diego, CA). Unconjugated mouse anti-CD162 receptor (CD162R; IgM, PEN5) was purchased from Beckman Coulter (Miami, FL). Isotype-matched, nonreactive mAb were obtained from BD Biosciences PharMingen. Cy5-conjugated Fab fragments of affinity-purified goat anti-mouse IgM, H chain-specific antibodies were obtained from Jackson ImmunoResearch Laboratories. Affinity-purified, FITC-conjugated goat anti-rat Ig and goat anti-human Ig and PE-conjugated goat anti-mouse Ig antibodies [F(ab')2 fragments] were obtained from Tago Inc. (Burlingame, CA). Peroxidase-conjugated goat anti-rat IgM and goat anti-mouse IgM and IgG were purchased from Pierce Chemical (Rockford, IL). Cy-chrome-conjugated anti-CD3 mAb was purchased from BD Biosciences PharMingen. PE-conjugated anti-CD56 (anti-Leu19), anti-CD16 (anti-Leu11c), anti-CD3 (anti-Leu4), anti-CD14 (anti-LeuM3), anti-CD19 (anti-Leu12), and anti-CD25 (anti-IL-2R) and FITC-conjugated anti-CD3 (anti-Leu4), anti-CD16 (anti-Leu11a), anti-CD14 (anti-LeuM3), anti-CD19 (anti-Leu12), anti-CD57 (human NK-1), anti-CD161, anti-human leukocyte antigen (HLA)-DR, anti-transferrin receptor (CD71), and anti-CD69 mAb were procured from BD Biosciences. FITC-conjugated anti-IL-2Rß mAb was purchased from Endogen Inc. (Boston, MA). Unconjugated and PE-conjugated anti-CD18, ant-CD11a, anti-intercellular adhesion molecule-1 (ICAM-1), anti-CD29, and anti-CD49f [very late antigen (VLA)-6] were obtained from AMAC (Westbrook, ME). Anti-CD2, anti-CD11b, and anti-CD11c mAb were purchased from BD Biosciences PharMingen. Anti-CD49b (VLA-2), anti-CD49c (VLA-3), anti-CD49d (VLA-4), and anti-CD49e (VLA-5) mAb were purchased from Telios (San Diego, CA). Anti-lymphocyte function-associated antigen-3 (HB205) and anti-CD16 (B73.1, Leu11c) mAb were generated in our laboratory as culture supernatants of the corresponding hybridomas (American Type Culture Collection, Manassas, VA).
Cell lines
Primary human umbilical vein endothelial cells (HUVEC, lot #30396) and cell systems (CS)-C endothelial cell growth medium system were obtained from Cell Systems (Kirkland, WA). HUVEC were cultured as recommended by the vendor. Briefly, HUVEC were resuspended in CS-1.0 growth medium and plated in T25 flasks (Costar, Cambridge, MA) or in 48-well plates (Costar). The tissue-culture surfaces were precoated with AF-1.0 HUVEC attachment medium. Cell suspensions of HUVEC were obtained by a brief trypsinization of fully confluent monolayers and used for further passage or for experiments.
K562, human myeloid leukemia, Daudi, human Burkitt’s lymphoma, and P815 mouse mastocytoma cell lines were cultured as cell suspensions in RPMI-1640 medium, supplemented with 10% (v/v) fetal calf serum (FCS; RPMI-10% FCS), L-glutamine, antibiotics, and HEPES buffer (Life Technologies, Long Island, NY).
Fresh human tissues
Peripheral blood buffy coats were obtained from normal blood donor volunteers via the Pittsburgh Central Blood Bank (PA). The buffy coats served as a source of peripheral blood mononuclear leukocytes (PBMNL). Various normal tissues, including spleen, tonsil, liver, lung, and intestine of normal individuals, noninvolved buccal mucosa, and noninvolved lymph nodes of head and neck cancer patients, and cancer tissues from various cancer patients were collected and processed using the Institutional Review Board-approved Informed Consent and Tissue Collection Protocol (#991205) and provided to the investigators by the University of Pittsburgh Cancer Institute Tissue Procurement Facility and University of Pittsburgh Department of Pathology (PA).
Separation of peripheral blood and tissue resident mononuclear leukocytes
PBMNL were purified from normal blood buffy coats by centrifugation on Ficoll-Hypaque density gradients. Suspensions of lymphocytes from spleen, tonsils, and lymph nodes were obtained by mechanical disaggregation of the lymphoid tissues followed by centrifugation on Ficoll-Hypaque gradients. Nonparenchymal, mononuclear leukocytes from liver, buccal mucosa, and cancer lesions were isolated and partially purified by preparation of single-cell suspensions using a collagenase treatment of the tissues, followed by their centrifugation on two-layer (70%/100%) Ficoll-Hypaque gradients, as described previously [36
].
Purification of PBMNL
NK cells, T cells, B cells, and monocytes were purified from normal human PBMNC using a modification of the negative immunoselection technique with antibody-coated magnetic beads obtained from Advanced Magnetics (Boston, MA) as described previously [37
] and were consistently 
92% pure populations of CD5–CD3–T cell receptor (TCR)-ß–CD14–CD19–CD56+CD16+, CD5+CD3+TCR-ß+CD14–CD19–CD56–CD16–, CD19+, and CD14+ leukocytes, respectively.
Antibody/complement-mediated elimination of ANK-1+ NK cells
Purified NK cells were preincubated on ice with anti-ANK-1 mAb (2 µg/106 cells) for 30 min and washed two times with RPMI 1640. The antibody-coated NK cells were resuspended in the 1/20 (v/v) complement solution in RPMI-10% FCS and incubated for 1 h at 37°C. Following this incubation, the number of lysed cells was determined by trypan blue dye exclusion assay, and cells were washed three times in RPMI-10% FCS.
Selection of A-NK and NA-NK cells
Purified NK cells (1x106/ml) were resuspended in RPMI-1640 medium supplemented with 10% human AB serum (NABI, Miami, FL), 2 mM L-glutamine, 50 U/ml penicillin, 50 µg/ml streptomycin, 250 µg/ml Fungizone, and 25 mM HEPES buffer. To block activation-induced protein synthesis and cell differentiation, a portion of NK cells was preincubated in the presence of 10 µg/ml Cx at 37°C for 1 h. Next, treated or untreated NK cells (10x106) were mixed with 22 nM IL-2, seeded in T75 tissue-culture flasks, and incubated at 37°C for 1 h. During this incubation, a portion of NK cells became A-NK cells. NA-NK cells were separated from A-NK cells by decanting the conditioned culture medium and floating NK (NA-NK) cells and by washing A-NK cells five times with warm (37°C) RPMI-1640 medium containing 2% FCS (RPMI-2% FCS) to eliminate residual NA-NK cells. A-NK cells were detached by incubation in Ca2+- and Mg2+-free phosphate-buffered saline (PBS) at 4°C for 30 min. Following their collection, A-NK and NA-NK cells were washed three times in RPMI-10% FCS.
Sorting of ANK-1+ and ANK-1– NK cells
Fresh, purified NK cells were suspended in RPMI-2% FCS, mixed with FITC-conjugated anti-ANK-1 mAb (5 µg/106 cells), and incubated on ice for 45 min to label ANK-1+ cells. The NK cells were then washed twice with Hanks’ solution supplemented with 2% FCS, resuspended in the same medium (107 cells/ml), and sorted in a FACStar Plus cytometer (Becton Dickinson, San Jose, CA). Sorted ANK-1+ and ANK-1– NK cells were >98% pure and viable populations.
Culture of A-NK cells and NA-NK cells
Purified NK cells were resuspended in RPMI-1640 medium supplemented with 10% AB serum and 22 nM IL-2, seeded in T75 tissue-culture flasks (Falcon, BD Labware, Franklin Lakes, NJ), and incubated for 5 h. After this initial incubation, A-NK and NA-NK cells were separated and continued to culture for 14 days in the presence of 22 nM IL-2, as described [23
].
Flow cytometry
Fresh PBMNL, purified NK cells, sorted NK cells, T cells, B cells, and monocytes or tissue mononuclear leukocytes were suspended in ice-cold PBS containing 0.1% sodium azide and 1% FCS (PBS-AF). For direct labeling, the cells (0.2x106/0.1 ml) were incubated on ice for 30 min with the indicated fluorochrome-conjugated mAb. For indirect labeling, the cells were incubated with unlabeled, primary mAb, washed twice with PBS-AF, and then incubated with fluorochrome-conjugated polyclonal antibodies. Negative controls were cells incubated without antibodies or with isotype-matched, fluorochrome-conjugated, nonreactive mAb or with isotype-matched, nonreactive mAb followed by the secondary fluorochrome-conjugated antibodies. After these incubations, the cells were washed twice with PBS-AF, fixed with 1% (w/v) paraformaldehyde/PBS solution, and analyzed by flow cytometry, as described previously [15
, 23
]. One-, two-, or three-color flow cytometry was performed on a Coulter Epics XL flow cytometer (Beckman Coulter), and analysis and presentation of data were executed using the EXPO32 program (Beckman Coulter).
Cell adherence assays
Forty eight-well plates (Falcon, BD Labware) were prepared for the assays by coating with HUVEC, fibronectin, laminin, collagen I, collagen IV, or BSA. Confluent monolayers of HUVEC were obtained, as described above. Coating of plastic surfaces in the wells with CAMs or BSA was achieved by adding into the wells the PBS solutions of the proteins (2 µg or 2 mg, respectively/200 µl/well) and by incubating the plates for 18 h at 4°C. The CAM-coated surfaces were additionally treated by 1 h incubation at 37°C in the presence of BSA solution (10 mg/ml). Following these incubations, wells were washed three times with PBS and twice with RPMI 1640, supplemented with 10% human AB serum. Next, purified, unsorted, or sorted NK cells were resuspended (2.5x104 cells/200 µl) in the same medium, supplemented with 22 nM IL-2, added to the wells, and incubated for 1 h at 37°C. After this incubation, NA-NK cells were removed, wells were washed three times with warm (37°C) RPMI-2% FCS, and A-NK cells were counted in seven different microscopic fields using an inverted microscope, an ocular grid, and 200x magnification. The assays were performed in triplicate.
Blocking of NK cell adherence by antibodies against adhesion molecules
Purified NK cells were preincubated for 2 h at 0°C in the presence of 10 µg/ml isotype control nonreactive mAb or various blocking mAb specific for CAMs. Following this treatment, NK cells were resuspended in RPMI-1640 medium containing 10% human AB serum and 22 nM IL-2, seeded in 48-well plates, and incubated at 37°C for 1 h. The rest of the assay was performed as described above for cell adherence assays.
Cytotoxicity assays
Standard 4 h 51Cr release cytotoxicity assays were performed in four different effector:target (E:T) ratios and in triplicate to measure NK activity against K562 cell targets and LAK activity against Daudi cell targets and redirected killing against P815 cell targets in the presence of anti-CD16 (B73.1, 0.1 µg/ml) mAb, as described previously [15
, 23
].
Immunohistochemistry
Fragments of normal spleens, lymph nodes, tonsils, intestines, lungs, and livers were fixed with 4% paraformaldehyde at 4°C for 4 h, dehydrated in ethanol, embedded in paraffin, and sectioned on a microtome. The tissue sections (5 µ) were deparaffinized with Histoclear, rehydrated, and incubated with a 0.4% pepsin solution for 10 min at 37°C to unmask antigenic epitopes. The tissue sections were blocked with 5% normal goat serum (Sigma Chemical Co.) for 40 min and incubated with anti-ANK-1 mAb diluted in PBS solution of 0.1% Triton X-100 (v/v) for 2 h at room temperature. The tissue sections were washed three times with PBS and incubated with biotinylated goat anti-rat IgM secondary antibodies (Vector, Burlingame, CA, 1/400 dilution) for 2 h, washed three times with PBS, and incubated with avidin-biotin complex (Elite ABC kit, Vector) for 2 h at room temperature. Color reaction was developed using 3-amino-9-ethyl carbazole. The tissue sections were counterstained at room temperature with Mayer’s hematoxylin (Sigma Chemical Co.) for 4 min and rinsed with tap water. The slides were mounted and visually analyzed, and images were collected using an Olympus light microscope.
Three-color immunohistofluorescence
Deparaffinized and hydrated lymph node sections were treated with 10 mM citrate buffer, pH 6.0 (antigen-retrieval solution), for 30 min in a steamer, cooled for 10 min at room temperature, and rinsed with distilled water for 10 min. The sections were then washed six times with PBS solution containing 0.5% BSA and 0.15% glycine, blocked with 5% normal goat serum (Sigma Chemical Co.) for 40 min, washed three times, incubated with rabbit anti-human CD3 antibodies (Dako Corp., Carpenteria, CA) and rat ant-mouse ANK-1 mAb for 1 h at room temperature, and washed three times. The sections were then incubated with CY3-conjugated goat anti-rabbit (Jackson ImmunoResearch Laboratories) and Alexa 488-conjugated goat anti-rat (Molecular Probes, Eugene, OR) antibodies for 1 h at room temperature and washed six times. Following this staining, the sections were incubated with Dapi solution (Sigma Chemical Co.) for 30 s, washed, and mounted. Next, three-color fluorescence analyses of the stained lymph node sections were performed, and images were collected using Olympus BX-51 fluorescent microscope and magnifier image acquisition software.
Preparation of NK cell lysates
Lysates were prepared by resuspending 107-purified NK cells in 1 ml ice-cold lysing buffer containing 10 mM EDTA, 4 mM phenylmethylsulfonyl fluoride, 80 µM antipain, 110 µM leupeptin, 60 µM pepstatin, 36 µg aprotinin (Sigma Chemical Co.), and 1% (v/v) Triton X-100 in PBS and incubating the suspension on ice for 30 min. Soluble material was separated from cell nuclei and cell debris by centrifugation at 16,000 g for 1 h at 4°C. The resulted soluble fraction was collected, aliquoted, and stored at –20°C.
Affinity purification
NK cell lysates were diluted with PBS solution containing 0.1% Triton X-100 and chromatographed at 4°C on a column of Tresyl-activated Sepharose (Pharmacia, Uppsala, Sweden) coupled with mAb ANK-1. Elution of bound material was accomplished using 0.5 M diethylamine, pH 11.5. The eluate was immediately adjusted to pH 7.6 and dialyzed against PBS.
Immunoprecipitation
Aliquots (100 µl) of NK cell lysate were precleared twice with 25 µl washed and packed Protein G-Sepharose 4 Fast Flow beads (Amersham Bioscience Co., Piscataway, NJ) by incubation on ice for 1 h, followed by centrifugation at 600 g for 5 min at 4°C. Precleared lysates were incubated with 10 µg antibody for 4 h at 4°C. Thus, formed immune complexes were collected by incubation and gentle mixing of lysates with 50 µl Protein G-Sepharose beads overnight at 4°C. After that, the beads with bound immune complexes were separated from the rest of lysates and washed four times in lysing buffer by centrifugation at 600 g for 5 min at 4°C.
Western blotting
Untreated lysates, affinity-purified ANK-1 antigen, lysates cleared with antibodies, beads with bound immunoprecipitates, and affinity-purified mouse brain N-CAM [22
, 23
] were mixed with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer without ß-mercaptoethanol in a 1:1 ratio, boiled for 3 min, and centrifuged. Antigens in the resulted supernatants were separated by SDS-PAGE on precast 7.5% linear or 4–15% gradient polyacrylamide gels (Bio-Rad Laboratories, Hercules, CA) and transferred onto a polyvinyl difluoride Immobilon-P (Millipore, Bedford, MA) membrane for 6 h at 350 mA on ice. The membranes were fixed with methanol for 30 s and air-dried for 15 min. Immunostaining was performed at room temperature in two steps. First, the membranes were incubated with solutions of primary ANK-1, 12F8, or isotype-matched control antibodies in Tris-buffered saline (TBS) solution containing 5% nonfat dry milk and 0.05% Tween (TBS/Tween) at room temperature for 1 h. The membranes were washed five times with TBS/Tween and incubated with peroxidase-conjugated goat anti-rat IgM antibodies (Pierce Chemical, 1/100,000 dilution) at room temperature for 1 h. The membranes were washed again and incubated with the enhanced chemiluminescence substrate (SuperSignal West Femto, Pierce Chemical). The antigens were then visualized by exposing the membranes to Kodak BioMax MR films (Rochester, NY).
Statistical analyses
Statistical analyses of the results were performed using the exact Wilcoxon-Mann-Whitney test, with or without stratification as needed. P values were adjusted by the step-down Bonferroni method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
250 kD molecules in the crude NK cell lysates, forming a single band (Fig. 2A
). Similar results were obtained with mAb 12F8. However, the band of the molecules reactive with mAb 12F8 had higher density and was broader than that reactive with mAb ANK-1, indicating that the 12F8 molecules are more abundant and more heterogeneous than the ANK-1 molecules on NK cells. More importantly, mAb 12F8 (Fig. 2B)
and polyclonal rabbit anti-mouse N-CAM antiserum (data not shown) specifically reacted with the affinity-purified ANK-1 molecules derived from NK cell lysates. These findings suggest that mAb ANK-1 specifically reacts with N-CAM expressed by human fresh NK cells.
|
Phenotype of ANK-1+ NK cells
To determine whether ANK-1+ NK cells are pre-A-NK cells, we purified fresh blood NK cells, sorted ANK-1+ and ANK-1– NK cells, and assessed these cells by flow cytometry for the coexpression of ANK-1 with CD56, CD16, IL-2Rs, CAMs, and HLA-DR, as these molecules may be expressed differentially by different subsets of NK cells or NK cells at different stages of maturation or activation [1
2
3
]. We found that ANK-1+ NK cells, in contrast to ANK-1– NK cells, expressed higher levels of several CAMs, including the ß1 integrins CD29, CD49d, and CD49e; the ß2 integrins CD18, CD11a, CD11b, and CD11c; and the Ig superfamily members CD54 and CD2. In addition, ANK-1+ NK cells expressed higher levels of CD122 (IL-2Rß). In contrast, ANK-1+ NK cells expressed lower levels of CD56 and CD16 than ANK-1– NK cells and did not express HLA-DR, CD25 (IL-2R; Table 1
), transferrin receptor, or CD69 (data not shown). These findings show that ANK-1+ and ANK-1– NK cells are phenotypically different. They also indicate that the phenotypic profiles of ANK-1+ and ANK-1– NK cells are similar to that of A-NK and NA-NK cells, respectively [9
, 21
]. These data suggest that ANK-1+ NK cells are pre-A-NK cells.
|
To confirm the possibility that ANK-1+ NK cells are mature NK cells, we examined fresh, purified, unsorted peripheral blood NK cells by three-color flow cytometry for the coexpression of ANK-1 and CD56 with CD161 (NKR-P1) or CD162R (PEN5). CD161 has been shown previously to be a marker of immature human NK cells, if expressed in the absence of CD56, or mature NK cells, if coexpressed with CD56 [32
]. CD162R (PEN5) is the mucin-like P-selectin glycoprotein ligand-1 selectively expressed on most mature CD56dimCD16+ NK cells [38
]. We consistently found that ANK-1+CD56dim NK cells coexpressed CD161 and that 27% of CD56dimCD162R+ NK cells coexpressed ANK-1 (data not shown). These findings further support the conclusion that ANK-1+ NK cells are mature NK cells.
Functions of ANK-1+ NK cells
A hallmark of pre-A-NK cells is their ability to respond rapidly to IL-2 by adherence to solid support surfaces. The finding of increased expression levels of integrins on ANK-1+ NK cells indicated that this subset of NK cells might be poised to respond rapidly to IL-2 by developing adhesiveness and therefore, might be pre-A-NK cells. To test this possibility, we compared IL-2-induced adhesiveness of sorted ANK-1+ and ANK-1– NK cells. We found that ANK-1+ NK cells responded rapidly to IL-2 by development of adhesiveness to HUVEC, the extracellular matrix (ECM) proteins fibronectin, laminin, and collagen, or to uncoated plastic surfaces, but not to BSA-coated plastic surfaces. In contrast to ANK-1+ NK cells, ANK-1– NK cells did not demonstrate the adhesion (Fig. 3A
). These findings show that ANK-1+ NK cells but not ANK-1– NK cells are constitutively endowed with the IL-2-inducible adhesiveness and further indicate that ANK-1+ NK cells are pre-A-NK cells.
|
Next, we examined cytotoxic activities of the sorted ANK-1+ and ANK-1– NK cells. We found that ANK-1+ and ANK-1– NK cells mediated direct NK cytotoxicity and redirected ADCC via FcRIII (Fig. 4
). However, although the two subsets of NK cells had similar NK cytotoxicity against K562 leukemia cell targets (Fig. 4A)
, ANK-1+ NK cells were significantly less capable of mediating redirected ADCC than ANK-1– NK cells (Fig. 4B)
. In addition, cytotoxic activities of A-NK cells and NA-NK cells selected by 1 h IL-2-induced adherence in the presence of Cx were consistently found in two different experiments to be similar to those of sorted ANK-1+ and ANK-1– NK cells, respectively (data not shown). These findings again suggest that ANK-1+ NK cells are a functionally distinct subpopulation of mature, nonactivated NK cells and that they are pre-A-NK cells.
|
24% of NK cells. At the same time, 76% of the NK cells able to adhere rapidly to the plastic surface after induction with 22 nM IL-2 was depleted. Following depletion of ANK-1+ NK cells, the remaining NK cells showed a markedly reduced ability to generate A-NK cells with LAK activity. Specifically, A-NK cells obtained after depletion of ANK-1+ NK cells expanded eightfold less and generated 69-fold less LAK activity than A-NK cells obtained from untreated NK cells. These findings demonstrate that the generation of A-NK cells with LAK activity is an exclusive property of the subset of mature, resting NK cells expressing the ANK-1 epitope. These findings further support the conclusion that the ANK-1 epitope represents a marker of human pre-A-NK cells.
|
26% of CD56+ NK cells. In addition, we found that a small subset of 5% CD3+ T cells expressed ANK-1. In contrast, CD14+ monocytes and CD19+ B cells showed no expression of ANK-1. Additional studies of five different normal samples of fresh PBMNL using two-color flow cytometry with PE-conjugated mAb to CD3, CD56, CD14, and CD19 lineage markers and FITC-conjugated mAb ANK-1 revealed that peripheral blood also contained a small subpopulation (1.44±0.19%, mean ±SD%) of lineage marker-negative cells, which expressed the ANK-1 epitope. Therefore, in normal human peripheral blood, the ANK-1 epitope is expressed on pre-A-NK cells, a small subset of T cells, and a minute subpopulation of cells lacking lineage markers.
ANK-1+ cells in solid tissues
Previous studies have shown that the ability to infiltrate solid tumor tissues is one of the characteristics of A-NK cells [24
, 39
]. This particular function enables A-NK cells to make contact with and efficiently kill cancer cells in tissues and consequently, eliminate established metastases and tumors [24
, 40
]. Therefore, we examined whether pre-A-NK (ANK-1+ NK) cells also have the ability to infiltrate solid, normal tissues and tumors. Using single-color immunohistochemistry, we found that ANK-1+ cells were present in significant numbers in defined areas of normal tonsils, spleen, lymph nodes, intestine, lungs, and liver. In the lymphoid organs, numerous ANK-1+ cells were found in germinal centers (Fig. 5
), and a few of these cells were evident throughout the other regions (data not shown). ANK-1+ cells were also found in the liver attached to the sinusoid lining endothelial cells, in the lungs in the interstitium, and in the intestine in the mucosa lamina propria (data not shown).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous studies have shown that human NK cells express 180, 140, and 120 kD N-CAM isoforms, which in native state, are polysialylated and have 230 and/or 180 kD MW [29 30 31 ]. These studies have also indicated that although the 140- and 120-kD N-CAM isoforms are abundant, the 180-kD N-CAM isoform is expressed at low levels on human NK cells. In the present study, we defined the novel N-CAM epitope ANK-1 and showed that the epitope is a characteristic residue of a fraction of the native 250-kD polysialylated CD56 N-CAM, which was preferentially expressed by a subset of NK cells comprising 26% of the peripheral blood NK cell population. The large molecular size of ANK-1 N-CAM and its exclusive presence in a fraction of NK cell N-CAM and its expression by only a small proportion of NK cells indicate that ANK-1 N-CAM might be the 180-kD N-CAM isoform. In addition, we found that the ANK-1 epitope is coexpressed with CD56 on a small subset of T cells but could also be expressed alone in the absence of the conventional CD56 epitope and CD3, by a discrete subpopulation of lymphocytes. These findings indicate that the ANK-1 epitope may also be expressed by the molecules that are different from CD56 N-CAM. It is possible that these molecules are N-CAM isoforms that do not express the conventional CD56 epitope or molecules different from N-CAM. These possibilities are currently under investigation in our laboratories.
Pre-A-NK cells were defined in fresh populations of purified CD3–CD56+ peripheral blood NK cells and were found to express typical markers of mature NK cells, including CD56, CD16, and CD161, and to lack CD25 and activation markers, such as CD69, transferrin receptor, and HLA-DR. In addition, pre-A-NK cells showed the ability to mediate spontaneously, without previous activation, direct lysis of the prototypical NK cell target K562 myeloid leukemia cells but not of the NK-resistant P815 mouse mastocytoma cell targets. These findings show that pre-A-NK cells represent a subset of mature, resting NK cells.
We determined that pre-A-NK cells, in contrast to the rest of fresh NK cells, exhibit two biologically important and inter-related, specific characteristics, i.e., increased expression of multiple integrins and a rapid response to IL-2 by developing integrin-mediated adherence to endothelial cells and ECM. These findings indicate that pre-A-NK cells might be constitutively able to bind rapidly to capillary endothelial cells and transmigrate into solid tissues. The inducible adhesiveness not only clearly distinguishes pre-A-NK cells from the other NK cells but may also represent a critical prerequisite for their transendothelial migration and infiltration of normal tissues, sites of immune responses, and tumors. Probably related to these functions, pre-A-NK cells were found to accumulate preferentially in solid, normal tissues such as lymph nodes, spleen, tonsils, liver, lungs, and intestinal and buccal mucosa, as well as in tumors.
Two other inter-related characteristics were demonstrated in pre-A-NK cells, i.e., lower levels of expression of FcRIII (CD16) and decreased levels of reverse ADCC. Previous studies have determined that human peripheral blood NK cells include the CD56dimCD16bright IL-2Rß+c-kit+ subset, which weakly responds to nanomolar concentrations of IL-2 and is a low producer of cytokines but mediates high levels of ADCC and NK cytolysis, and the CD56brightCD16– IL-2Rß+c-kit– subset, which vigorously responds to picomolar concentrations of IL-2 and is a high producer of cytokines but is unable to mediate ADCC or natural cytotoxicity [16
, 18
, 21
, 22
]. Our study shows that CD56dim NK cells are heterogeneous and consisted of the previously described CD16bright NK cells and the CD16dim pre-A-NK cells described herein, which exhibit high and low levels of ADCC, respectively. This conclusion is supported by the previous demonstration that most of the resident NK cells in the liver and secondary lymphoid tissues are CD16dim/or– [48
, 49
] and our findings that most of NK cells in these tissues are phenotypically pre-A-NK cells.
CD57+ lymphocytes, which could be NK cells or T cells [17 ], have been found by single-color immunohistochemistry to be present in secondary lymphoid organs and are mostly located in the germinal centers of secondary lymphoid follicles [50 ]. Our studies, using single-color immunohistochemistry and three-color immunohistofluorescence, demonstrated that in secondary lymphoid tissues, ANK-1+ lymphocytes had a very similar distribution to that of CD57+ lymphocytes and that they were preferentially located in germinal centers, were mostly CD3–, and were positioned in a close proximity to the few T cells present there. More detailed analyses with two- and three-color flow cytometry demonstrated that CD3–CD56dimANK-1+ lymphocytes, which were likely pre-A-NK cells, were present in substantial numbers and were the major NK cells in secondary lymphoid organs. In contrast, very few if any CD3–CD56brightANK-1– NK cells were found in these tissues. Thus, our studies show that secondary lymphoid tissues contain mostly CD56dimANK-1+ NK cells, which preferentially localize into germinal centers. Recent studies have indicated that reactive, secondary lymphoid tissues, in particular, lymph nodes, contain CD56brightCD16–KIR–natural cytotoxicity receptor–perforin–, noncytotoxic NK cells, which appear to be located in the PTAs of lymph nodes [49 , 51 ]. However, in contrast to peripheral blood CD56bright NK cells, the lymph node CD56bright NK cells express KIRs CD16 and perforin and become cytotoxic upon stimulation with IL-2 [21 , 22 , 49 ]. In addition, cytokines such as IL-2 and IL-12, which may be produced in reactive peripheral lymphoid tissues and are potent activators of NK cells, induce large increases in the expression of CD56 in CD56dim and CD56bright NK cells [23 , 52 ]. Therefore, the high expression levels of CD56, which have been observed in reactive lymph nodes, may be associated with cytokine-activated CD56dim NK cells but not with quiescent CD56bright NK cells, as has been suggested [21 , 22 , 49 , 51 ]. As NK cells have been shown to collaborate with dendritic cells and other immune cells and to regulate immune reactions via production of proinflammatory and immunoregulatory cytokines [1 2 3 , 12 13 14 , 21 , 22 ], the preferential distribution of CD3–CD56dimANK-1+ NK (pre-A-NK) cells in germinal centers of peripheral lymphoid tissues may indicate their active involvement and particular role in immune responses.
CD57+ lymphocytes have also been found to infiltrate human tumors. Although present in low numbers, their frequency has been found to correlate positively with prognosis of cancer patients [47 ]. Again, these studies have not defined whether the CD57+ TILs are NK cells and/or T cells. Using two- and three-color flow cytometry, we consistently found that TILs obtained from various human tumors contained a substantial population of NK cells that had mostly a CD3–CD56dimANK-1+ phenotype, which was consistent with pre-A-NK cells. In contrast, TILs contained a minute subset of lymphocytes, which were CD3+CD56dim. These studies indicate that tumors, similar to secondary lymphoid tissues, are preferentially infiltrated with pre-A-NK cells. The tumor-infiltrating NK cells might be able directly and/or via activation of other tumor-infiltrating immune cells to contribute to the immune control of tumor growth.
In conclusion, we phenotypically and functionally defined pre-A-NK cells as a distinct subpopulation of quiescent, mature peripheral blood NK cells, which differ from the rest of NK cells by the expression of the ANK-1 N-CAM epitope, high levels of integrins and IL-2Rß, and low levels of CD16; by the abilities to mediate NK activity effectively but ineffectively reverse ADCC; by the development of adhesiveness, expansion, and LAK activity upon activation with IL-2; and by the preferential accumulation in solid, normal tissues, germinal centers of secondary lymphoid tissues, and in tumors. The biological and clinical importance of pre-A-NK cells awaits elucidation in future studies.
|
ACKNOWLEDGEMENTS |
|---|
Received August 6, 2002; revised August 11, 2004; accepted August 16, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
