IMMUNOLOGICAL INVESTIGATIONS, 21(1),
11-23 (1992)
PRODUCTION AND CHARACTERIZATION OF A NOVEL MONOCLONAL ANTIBODY INHIBITORY FOR MURME NATURAL KILLER CELL ACI'IVITY D.W. Hoskin' and J.C. Rode? Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
' Department of Microbiology, Dalhousie University, Halifax, Nova Scotia B3H 4H7
* Division of Molecular Immunology and Neurobiology, Mount Sinai Hospital Research Institute, Toronto, Ontario M5G 1x5
ABSTRACT Natural killer (NK) cells are considered to play an important role in tumor surveillance. The killing of tumor target cells by NK cells is the result of a complex series of sequential binding, signal processing and lytic events. However, the mechanism which NK cells use to recognize tumor targets is poorly understood. To further study the cellsurface molecules involved in tumor recognition, we immunized rats against cloned murine T cells with NK activity (DBA/2.1) and generated rat-mouse hybridomas which were screened for the ability to block lytic activity of DBA/2.1 effector cells. Culture supernatants from one IgM-producing hybridoma, designated SlC4, were found to consistently inhibit DBA12.1-mediated lysis of YAC-1 target cells. Endogenous splenic NK activity was also diminished in the presence of S1C4 monoclonal antibody (mAb) while alloantigen-specific cytotoxic T lymphocyte (CTL) activity was not affected. S1C4 mAb appears to react with effector cell-surface structures involved in the recognition/adhesion phase of NK activity since pretreatment of effector cells with mAb S1C4 inhibits their ability to bind to YAC-1 target cells. ELISA studies revealed that the S1C4 antigen is expressed by a range of lymphoid cell lines, as well as by DBA/2.1 cells and fresh splenic NK cells. S1C4 mAb were shown to react with 22, 24, 30, and 46 kiloDalton (kDa) DBA/2.1 cell membrane components on immunoblotsperformed under reducing conditions. These structures do not correspond to any known recognition/adhesion molecules, suggesting that mAb S1C4 defines novel cell membrane components involved in NK cell function.
INTRODUCTION Natural killer (NK) cells are a subpopulation of lymphocytes characterized by their ability to recognize and lyse a wide range of tumor cells in an MHC-unrestricted manner without the need for prior antigen priming (1,2). The lytic effect of NK cells on malignant cells in vitrQ (1,2) and the positive correlation which exists between the level of endogenous
NK activity in mice and in v i resistance ~ ~ to certain tumor cell lines (3) lend support to the hypothesis that NK cells serve as an important barrier against the development and metastatic spread of tumor cells&triyp (1.2). NK cells are predominantly of large granularlymphocyte morphology (43) and, in the mouse, express the asialo-GM1, NK1 or NK2, and CD45 markers (1,2). The majority of cells with NK activity lack T cell-associated 11 Copyright 0 1992 by Marcel Dekka, Inc.
12
HOSKIN AND RODER
antigens such as CD3 and CD8 (1,6), although a subpopulation of NK cells are Thy-1 positive (7). Thus,the lineage of NK cells has remained obscure and is still controversial (8).
Tumor cell lysis by NK cells is the consequence of a complex sequence of cellular interactions involving 1) recognition and binding of effector cells to target cells, 2) Cat'dependent programming of the target cell for lysis, and 3) killer cell-independent cytolysis
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
of the target cell (9). NK cell-surface CD2 (lo), LFA-1 adhesion molecules (ll), very late activation antigen-like receptors (12), and CD45T200 glycoprotein (13) have all been implicated in the adhesion phase of MHC-independent cytotoxicity while neutral serine proteases and perforin/cytolysin molecules contained in the cytoplasmic granules of NK cells are thought to be important in the later stages of target cell destruction (14). However, the precise mechanism by which NK cells recognize tumor target cells is still illdefined. It is evident that NK cells do not utilize T cell receptor (TcR) structures in tumor target recognition since TcR genes are not rearranged and the TcR/CD3 complex is not expressed by endogenous NK cells (15,16,17). The fact that clonal NK lines show multiple target specificity suggests that the NK receptor (NKR) is not a clonotypic recognition structure (18). Thus, experimental evidence currently favors a widely distributed NKassociated target recognition mechanism of remarkably broad specificity (19).
In the present study we have utilized hybridoma technology to identify cell membrane components involved in natural cell-mediated cytotoxicity. A panel of monoclonal antibodies (mAb) prepared against the cloned DBAI2.1 NK-like T cell line were screened and selected for the ability to block killing of NK-sensitive YAC-1 target cells. This report describes the generation and characterization of hybridoma S1C4 which produces mAb of IgM isotype with the capacity to inhibit NK activity by interfering with effector-targetcell conjugate formation. Our results indicate that mAb S1C4 does not react with any known antigen receptor or cell adhesion molecule and m a y therefore define novel cell-surface structures involved in tumor recognition/adhesion by NK cells. MATERIALS AND METHODS Animals: Six- to eight-week-old male DBA/2 (H-2d), CBA/J (H-2k)and C3H/Hej (H-2k)mice were purchased from the Jackson Laboratory (Bar Harbour, ME). Young adult male Wistar rats were obtained from Charles River Canada (St. Constant, Que.). All animals were maintained on standard rodent chow and water in our facilities at the Mount Sinai Hospital Research Institute and Dalhousie University. Cell lines: The DBA/2.1 line of IL-2 dependent MHC-independent CTL (20) was generously donated by Dr. A. Lagarde (Toronto, Ont.). The A 2 O H L B cell lymphoma was kindly provided by Dr. A. Ochi (Toronto, Ont.) while the M1-A5 NC cell line was a gift from Dr. B. Pope (Halifax, N.S.). Hybridoma FD441.8 produces rat mAb directed against the alpha chain of murine LFA-1(21) and was obtained from the American Tissue Culture
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
NATURAL KILLER CELL ACTIVITY
13
Collection (Rockville, MD). All hybridomas and cell lines, with the exception of the DBA/2.1 and M1-AS lines, were maintained by continuousjn vitro passage in RPMI 1640 medium (Flow Laboratories, Mississauga, Ont.) supplemented with 5 x loJ M 2mercaptoethanol (Eastman-Kodak Co.,Rochester, NY),100 ug/ml streptomycin (Flow Labs), 100 U/ml penicillin (Flow Labs), 10 mM Lglutamine (Flow Labs) (hereafter referred to as complete medium) plus 10% heat-inactivated fetal calf serum (FCS)(Gibco/BRL, Burlington, Ontario). The DBA/2.1 and M1-A5 lines were propagated in complete RPMI 1640 medium plus 10% FCS supplemented with either 10% Con A conditioned medium, or 20% W H I - 3 culture supernatant, respectively. H v b r i d o m ' m : The procedure of somatic cell hybridization described by Kohler and Milstein (22) was employed to produce mAb with specificity for DBA/2.1 MHC-independent CTL. Briefly, Wistar rats were injected subcutaneously and intraperitoneally with 25 x 10s DBA/2.1 cells. The procedure was repeated twice at 2-week intervals. The animals were rested for one month and then inoculated intravenously with DBA/2.1 cells. Three days later spleens were removed aseptically and single cell suspensions were prepared. Next, l@rat lymphocytes were hybridized with 2 x lo7 cells of the nonsecreting thioguanine-resistantHGPRT negative Balb/c-derived mouse myeloma SP2/0 (ATCC). Cells were fused with 3350 mol. wt. 50% polyethylene glycol (Sigma Chemicals, St. Louis, MO). Hybridomas were maintained in complete RPMI 1640 medium plus 10% FCS and culture supernatants screened for inhibitory effects on DBA/2.1 lytic activity versus NK-sensitive YAC-1 target cells in a 4 hr 51Cr-releaseassay. Positive clones were subcloned by limiting dilution until stable mAb production was achieved. The isotype and concentration of mAb in culture supernatants was determined by ELISA. molytic 51~-release assay: The 5'Cr-release assay was performed using DBA/2.1 CI'L, fresh spleen cells, or alloantigen-specific CTL at various effector-to-target ratios in complete RPMI 1640 medium plus 10% FCS in "V-bottom microtiter plates (Flow Labs). NK-sensitive YAC-1 (H-2') and NK-resistant P815 (H-2d) target cells were labelled with N%51Cr0,. Blocking experiments were performed by pre-treating effector cells for 30 min with mAb (2 pg/ml) from anti-DBA/2.1 hybridoma SlC4, anti-LFA-1 hybridoma FD441.8, an irrelevant rat IgM-producing hybridoma, or control SP2/0 culture supernatants prior to addition of target cells. Plates were incubated for 4 hr at 3?C, centrifuged, and the supernatants collected and assayed for 51Cr-releasefrom lysed target cells with an LKB 1272 Clinigamma counter. Percent lysis was calculated according to the formula % lysis = E S x 100 M-S
-
where E = release from experimental samples, S = spontaneous release, and M = maximum release upon lysis with 10% Triton X-100. ConiuPate inhibition assay: The conjugate inhibition assay was performed by mixing 10' mouse spleen cells with an equal number of YAC-1 target cells in a 12 ml round
14
HOSKIN AND RODER
bottom tube. Next 1 pg of ammonium sulphate-precipitated mAb (SlC4 or an irrelevant
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
rat IgM) was added to the tube and the total volume adjusted to 0.5 ml with RPMI 1640 medium containing 5% FCS. Alternatively, in some experimentseffector cells and/or target cells were pre-incubated with mAb S1C4 for 1 hr at 4°C followed by extensive washing prior to use in order to eliminate unbound mAb from the assay. The tubes were centrifuged for 5 min at 700 rpm and incubated for 30 min at 37°C. Then the tubes were placed on ice and the conjugates suspended by mixing five times with a pipetteman set at 200 pl. Conjugated spleen cells and YAC-1 target cells were enumerated by microscopic observation in a hemacytometer. At least 200 spleen cells were counted for each determination. The percentage of spleen cells bound to target cells was determined by the formula [(number of effector cells bound to one or more target cells)/(total number of effector cells)] x 100. Cells were fixed to 96-well polyvinyl-chloride microtiter plates (Becton Dickinson and Co.,Lincoln Park, NJ.) with a 0.5% solution of glutaraldehyde (Sigma Chemical Co.) in PBS as described by Cole 4 (23). Nonspecific binding sites were
u:
blocked for at least 48 hr with 1%RIA-grade bovine serum albumin (BSA) in PBS. Plates were washed four times with PBS and 50 p1 of S1C4 culture supernatant (2 p g / d mAb) was added to each well. Negative controls consisted of RPMI 1640 medium and normal rat serum diluted 1/1O,OOO. Following a 90 min incubation at room temperature, the plates were washed four times with PBS and 50 p1 of a 1/1OOO dilution of horse radish peroxidaseconjugated goat anti-rat IgM (Daymar Laboratories, Toronto, Ont.) in 1% BSA/PBS was added to each well. The plates were incubated at room temperature for an additional 90 min, washed four times with PBS, and developed with 100 pl/well of substrate solution consisting of 2 mg/ml o-phenylenediamine(Sigma Chemicals Co.) and 0.015% H202 in 50 mM citrate buffer (PH 5.0). The plates were incubated in the dark for 30 min and the reaction stopped with 25 fi1 8N €-I$O,. Absorbance at 492 nm was determined on an automated ELISA plate reader (Flow Labs). The procedure followed was described by BioRad Laboratories (Richmond, CA). Briefly, cell membrane extracts prepared by extraction with 1% NP-40 in PBS with 1 mM phenylmethane sulfonylfluoride (Sigma Chemical Co.) were electrophoresed under reducing conditions on 10% SDS-PAGE slab gels. The gels were transblotted onto nitrocellulose paper which was then blocked with 5% Blotto for 1 hr. The nitrocellulose paper was incubated with ammonium sulphate-precipitated mAb S1C4 (1pg/ml) or an equivalent amount of anti-CD8(Lyt2.2) (rat IgM, Cedarlane, Hornby, Ont.) for 1 hr, washed with Tween-20-containingTris buffered saline (pH 8) and incubated with a 1/1Wdilution of '261-labelled sheep anti-rat immunoglobulin(Amersham Corp.) for 1hr. After washing, immunoblots were exposed to Kodak XAR5 autoradiographyfilm (Eastman Kodak Co.) with intensifying screens at -70°C. :-
15
NATURAL KILLER CELL A C T I V I T Y
RESULTS
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
. .. ' Generation of a mAb i n h h z y for NKactwly:Our strategy was to immunize rats with cells from the cloned DBA/2.1 T cell line (20) on the assumption that these IL2dependent CTL which exhibit NK-like MHC-unrestricted cytotoxicity employ NKR structures in the recognition of tumor target cells. DBA/2.1 cells display both T lymphocyte and NK cell characteristics, i.e., they rearrange and express TcR genes; and are CD3+, Thyl.2'; but are CDS; CD8- and express the NK-associated markers NK2 and asialoGM1; and are highly cytotoxic for NK-sensitive YAC-1 tumor target cells (24). Upon completion of the immunization protocol, immune rat spleen cells were prepared and fused with SP2/0 myeloma cells. The resulting hybridomas were plated at a dilution calculated to yield one clone per well. Once growth was observed the clones were immediately subcloned at 0.3 cells/well. As shown in Table I, of the 12 subclones tested, culture supernatants from hybridoma S1C4 consistently showed significant (pc 0.01) inhibitory activity on DBA/2.1mediated cytolysis of YAC-1 target cells. The culture supernatants of hybridoma S1C4 were analyzed by ELISA and found to contain approximately 2 pg/ml IgM antibody. Given that the DBA/2.1 cells used for immunizations are not true NK cells but are best described as NK-like Cm, it was important to determine whether mAb S1C4 also inhibited the function of fresh splenic NK cells. The data shown in Table I1 indicate that endogenous NK activity of fresh DBA/2 (expt. 1) or CBA/J (expt. 2) spleen cells is partially inhibited by the presence of mAb SlC4. The inhibitory effect of mAb S1C4 on NK activity is dose-dependent, decreasing to negligible levels at mAb concentrations below 0.25 pg/ml (data not shown). In contrast to anti-LFA-1 mAb which was used as a positive control and had inhibitory effects on both NK and CTL activity, mAb S1C4 had no effect on killing mediated by alloantigen-specificCTL. The addition of an irrelevant control rat IgM to the assay had little or no effect on either NK or Cll function. Conjugate inhibition assays were performed to determine whether mAb S1C4 acts by interfering with effector-target cell conjugate formation. As shown in Table III, expt. 1, the presence of mAb S1C4 results in a 45% inhibition of conjugate formation between fresh spleen cells and YAC-1 target cells. The inhibitory effect is mediated at the level of the effector cell since pretreatment of spleen cells with mAb S1C4 followed by washing effectivelyinhibited conjugate formation to the same extent as when mAb S1C4 was present throughout the assay (expt. 2). Pretreatment of target cells had no effect on killing. Cellular distribution of the S1C4 anti=: Next we employed a cellular ELISA to examine the reactivity of mAb S1C4 with various murine lymphoid and non-lymphoid cell populations. This assay has previously been shown to be at least as sensitive as immunoperoxidase staining and fluorescence-activated cell sorter analysis (23). Representative data are shown in Table IV. As expected, mAb S1C4 was found to react stronglywith the immunizingDBA/2.1 NK-like CTL line and with asialoGM1' splenocytes, as well as with unselected DBA/2 spleen cells. Binding of mAb S1C4 to spleen cells of
HOSKIN A N D RODER
16
TABLE I Inhibitory Effect of S1C4 Hybridoma Supernatant on NK Activity of DBA/2.1 Cells
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
Culture supernatants added to assay"
% Cytotoxicity E/T 5:l
% Inhibitionb
22 27 19 16 21 26 22 23 20 22 25 25 23
SP2/0 control S1B9 SIC2 S1C4
SlFl S1G7 S1H7 S2A8 S2E12 S2H12 S3C10 S3G6 S4G2
0 14 27c 5 0 0 0 9 0 0 0 0
a 25,000 DBA/2.1 effector cells per microtiter well were pre-incubated for 30 min at 37°C with a 35 dilution of culture supernatants from SP2/0 myeloma cells or anti-DBA/2.1 hybridomas, followed by the addition of 5,000 "Cr-labeled YAC-1 target cells per well to give an E/T ratio of 51. % Inhibition = % cvtotoxicity DBA/2.1 cells + hvbridoma supernatant X 100 % cytotoxicity DBA/2.1 cells plus SP 2/0 supernatant 'P
11-23 (1992)
PRODUCTION AND CHARACTERIZATION OF A NOVEL MONOCLONAL ANTIBODY INHIBITORY FOR MURME NATURAL KILLER CELL ACI'IVITY D.W. Hoskin' and J.C. Rode? Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
' Department of Microbiology, Dalhousie University, Halifax, Nova Scotia B3H 4H7
* Division of Molecular Immunology and Neurobiology, Mount Sinai Hospital Research Institute, Toronto, Ontario M5G 1x5
ABSTRACT Natural killer (NK) cells are considered to play an important role in tumor surveillance. The killing of tumor target cells by NK cells is the result of a complex series of sequential binding, signal processing and lytic events. However, the mechanism which NK cells use to recognize tumor targets is poorly understood. To further study the cellsurface molecules involved in tumor recognition, we immunized rats against cloned murine T cells with NK activity (DBA/2.1) and generated rat-mouse hybridomas which were screened for the ability to block lytic activity of DBA/2.1 effector cells. Culture supernatants from one IgM-producing hybridoma, designated SlC4, were found to consistently inhibit DBA12.1-mediated lysis of YAC-1 target cells. Endogenous splenic NK activity was also diminished in the presence of S1C4 monoclonal antibody (mAb) while alloantigen-specific cytotoxic T lymphocyte (CTL) activity was not affected. S1C4 mAb appears to react with effector cell-surface structures involved in the recognition/adhesion phase of NK activity since pretreatment of effector cells with mAb S1C4 inhibits their ability to bind to YAC-1 target cells. ELISA studies revealed that the S1C4 antigen is expressed by a range of lymphoid cell lines, as well as by DBA/2.1 cells and fresh splenic NK cells. S1C4 mAb were shown to react with 22, 24, 30, and 46 kiloDalton (kDa) DBA/2.1 cell membrane components on immunoblotsperformed under reducing conditions. These structures do not correspond to any known recognition/adhesion molecules, suggesting that mAb S1C4 defines novel cell membrane components involved in NK cell function.
INTRODUCTION Natural killer (NK) cells are a subpopulation of lymphocytes characterized by their ability to recognize and lyse a wide range of tumor cells in an MHC-unrestricted manner without the need for prior antigen priming (1,2). The lytic effect of NK cells on malignant cells in vitrQ (1,2) and the positive correlation which exists between the level of endogenous
NK activity in mice and in v i resistance ~ ~ to certain tumor cell lines (3) lend support to the hypothesis that NK cells serve as an important barrier against the development and metastatic spread of tumor cells&triyp (1.2). NK cells are predominantly of large granularlymphocyte morphology (43) and, in the mouse, express the asialo-GM1, NK1 or NK2, and CD45 markers (1,2). The majority of cells with NK activity lack T cell-associated 11 Copyright 0 1992 by Marcel Dekka, Inc.
12
HOSKIN AND RODER
antigens such as CD3 and CD8 (1,6), although a subpopulation of NK cells are Thy-1 positive (7). Thus,the lineage of NK cells has remained obscure and is still controversial (8).
Tumor cell lysis by NK cells is the consequence of a complex sequence of cellular interactions involving 1) recognition and binding of effector cells to target cells, 2) Cat'dependent programming of the target cell for lysis, and 3) killer cell-independent cytolysis
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
of the target cell (9). NK cell-surface CD2 (lo), LFA-1 adhesion molecules (ll), very late activation antigen-like receptors (12), and CD45T200 glycoprotein (13) have all been implicated in the adhesion phase of MHC-independent cytotoxicity while neutral serine proteases and perforin/cytolysin molecules contained in the cytoplasmic granules of NK cells are thought to be important in the later stages of target cell destruction (14). However, the precise mechanism by which NK cells recognize tumor target cells is still illdefined. It is evident that NK cells do not utilize T cell receptor (TcR) structures in tumor target recognition since TcR genes are not rearranged and the TcR/CD3 complex is not expressed by endogenous NK cells (15,16,17). The fact that clonal NK lines show multiple target specificity suggests that the NK receptor (NKR) is not a clonotypic recognition structure (18). Thus, experimental evidence currently favors a widely distributed NKassociated target recognition mechanism of remarkably broad specificity (19).
In the present study we have utilized hybridoma technology to identify cell membrane components involved in natural cell-mediated cytotoxicity. A panel of monoclonal antibodies (mAb) prepared against the cloned DBAI2.1 NK-like T cell line were screened and selected for the ability to block killing of NK-sensitive YAC-1 target cells. This report describes the generation and characterization of hybridoma S1C4 which produces mAb of IgM isotype with the capacity to inhibit NK activity by interfering with effector-targetcell conjugate formation. Our results indicate that mAb S1C4 does not react with any known antigen receptor or cell adhesion molecule and m a y therefore define novel cell-surface structures involved in tumor recognition/adhesion by NK cells. MATERIALS AND METHODS Animals: Six- to eight-week-old male DBA/2 (H-2d), CBA/J (H-2k)and C3H/Hej (H-2k)mice were purchased from the Jackson Laboratory (Bar Harbour, ME). Young adult male Wistar rats were obtained from Charles River Canada (St. Constant, Que.). All animals were maintained on standard rodent chow and water in our facilities at the Mount Sinai Hospital Research Institute and Dalhousie University. Cell lines: The DBA/2.1 line of IL-2 dependent MHC-independent CTL (20) was generously donated by Dr. A. Lagarde (Toronto, Ont.). The A 2 O H L B cell lymphoma was kindly provided by Dr. A. Ochi (Toronto, Ont.) while the M1-A5 NC cell line was a gift from Dr. B. Pope (Halifax, N.S.). Hybridoma FD441.8 produces rat mAb directed against the alpha chain of murine LFA-1(21) and was obtained from the American Tissue Culture
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
NATURAL KILLER CELL ACTIVITY
13
Collection (Rockville, MD). All hybridomas and cell lines, with the exception of the DBA/2.1 and M1-AS lines, were maintained by continuousjn vitro passage in RPMI 1640 medium (Flow Laboratories, Mississauga, Ont.) supplemented with 5 x loJ M 2mercaptoethanol (Eastman-Kodak Co.,Rochester, NY),100 ug/ml streptomycin (Flow Labs), 100 U/ml penicillin (Flow Labs), 10 mM Lglutamine (Flow Labs) (hereafter referred to as complete medium) plus 10% heat-inactivated fetal calf serum (FCS)(Gibco/BRL, Burlington, Ontario). The DBA/2.1 and M1-A5 lines were propagated in complete RPMI 1640 medium plus 10% FCS supplemented with either 10% Con A conditioned medium, or 20% W H I - 3 culture supernatant, respectively. H v b r i d o m ' m : The procedure of somatic cell hybridization described by Kohler and Milstein (22) was employed to produce mAb with specificity for DBA/2.1 MHC-independent CTL. Briefly, Wistar rats were injected subcutaneously and intraperitoneally with 25 x 10s DBA/2.1 cells. The procedure was repeated twice at 2-week intervals. The animals were rested for one month and then inoculated intravenously with DBA/2.1 cells. Three days later spleens were removed aseptically and single cell suspensions were prepared. Next, l@rat lymphocytes were hybridized with 2 x lo7 cells of the nonsecreting thioguanine-resistantHGPRT negative Balb/c-derived mouse myeloma SP2/0 (ATCC). Cells were fused with 3350 mol. wt. 50% polyethylene glycol (Sigma Chemicals, St. Louis, MO). Hybridomas were maintained in complete RPMI 1640 medium plus 10% FCS and culture supernatants screened for inhibitory effects on DBA/2.1 lytic activity versus NK-sensitive YAC-1 target cells in a 4 hr 51Cr-releaseassay. Positive clones were subcloned by limiting dilution until stable mAb production was achieved. The isotype and concentration of mAb in culture supernatants was determined by ELISA. molytic 51~-release assay: The 5'Cr-release assay was performed using DBA/2.1 CI'L, fresh spleen cells, or alloantigen-specific CTL at various effector-to-target ratios in complete RPMI 1640 medium plus 10% FCS in "V-bottom microtiter plates (Flow Labs). NK-sensitive YAC-1 (H-2') and NK-resistant P815 (H-2d) target cells were labelled with N%51Cr0,. Blocking experiments were performed by pre-treating effector cells for 30 min with mAb (2 pg/ml) from anti-DBA/2.1 hybridoma SlC4, anti-LFA-1 hybridoma FD441.8, an irrelevant rat IgM-producing hybridoma, or control SP2/0 culture supernatants prior to addition of target cells. Plates were incubated for 4 hr at 3?C, centrifuged, and the supernatants collected and assayed for 51Cr-releasefrom lysed target cells with an LKB 1272 Clinigamma counter. Percent lysis was calculated according to the formula % lysis = E S x 100 M-S
-
where E = release from experimental samples, S = spontaneous release, and M = maximum release upon lysis with 10% Triton X-100. ConiuPate inhibition assay: The conjugate inhibition assay was performed by mixing 10' mouse spleen cells with an equal number of YAC-1 target cells in a 12 ml round
14
HOSKIN AND RODER
bottom tube. Next 1 pg of ammonium sulphate-precipitated mAb (SlC4 or an irrelevant
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
rat IgM) was added to the tube and the total volume adjusted to 0.5 ml with RPMI 1640 medium containing 5% FCS. Alternatively, in some experimentseffector cells and/or target cells were pre-incubated with mAb S1C4 for 1 hr at 4°C followed by extensive washing prior to use in order to eliminate unbound mAb from the assay. The tubes were centrifuged for 5 min at 700 rpm and incubated for 30 min at 37°C. Then the tubes were placed on ice and the conjugates suspended by mixing five times with a pipetteman set at 200 pl. Conjugated spleen cells and YAC-1 target cells were enumerated by microscopic observation in a hemacytometer. At least 200 spleen cells were counted for each determination. The percentage of spleen cells bound to target cells was determined by the formula [(number of effector cells bound to one or more target cells)/(total number of effector cells)] x 100. Cells were fixed to 96-well polyvinyl-chloride microtiter plates (Becton Dickinson and Co.,Lincoln Park, NJ.) with a 0.5% solution of glutaraldehyde (Sigma Chemical Co.) in PBS as described by Cole 4 (23). Nonspecific binding sites were
u:
blocked for at least 48 hr with 1%RIA-grade bovine serum albumin (BSA) in PBS. Plates were washed four times with PBS and 50 p1 of S1C4 culture supernatant (2 p g / d mAb) was added to each well. Negative controls consisted of RPMI 1640 medium and normal rat serum diluted 1/1O,OOO. Following a 90 min incubation at room temperature, the plates were washed four times with PBS and 50 p1 of a 1/1OOO dilution of horse radish peroxidaseconjugated goat anti-rat IgM (Daymar Laboratories, Toronto, Ont.) in 1% BSA/PBS was added to each well. The plates were incubated at room temperature for an additional 90 min, washed four times with PBS, and developed with 100 pl/well of substrate solution consisting of 2 mg/ml o-phenylenediamine(Sigma Chemicals Co.) and 0.015% H202 in 50 mM citrate buffer (PH 5.0). The plates were incubated in the dark for 30 min and the reaction stopped with 25 fi1 8N €-I$O,. Absorbance at 492 nm was determined on an automated ELISA plate reader (Flow Labs). The procedure followed was described by BioRad Laboratories (Richmond, CA). Briefly, cell membrane extracts prepared by extraction with 1% NP-40 in PBS with 1 mM phenylmethane sulfonylfluoride (Sigma Chemical Co.) were electrophoresed under reducing conditions on 10% SDS-PAGE slab gels. The gels were transblotted onto nitrocellulose paper which was then blocked with 5% Blotto for 1 hr. The nitrocellulose paper was incubated with ammonium sulphate-precipitated mAb S1C4 (1pg/ml) or an equivalent amount of anti-CD8(Lyt2.2) (rat IgM, Cedarlane, Hornby, Ont.) for 1 hr, washed with Tween-20-containingTris buffered saline (pH 8) and incubated with a 1/1Wdilution of '261-labelled sheep anti-rat immunoglobulin(Amersham Corp.) for 1hr. After washing, immunoblots were exposed to Kodak XAR5 autoradiographyfilm (Eastman Kodak Co.) with intensifying screens at -70°C. :-
15
NATURAL KILLER CELL A C T I V I T Y
RESULTS
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
. .. ' Generation of a mAb i n h h z y for NKactwly:Our strategy was to immunize rats with cells from the cloned DBA/2.1 T cell line (20) on the assumption that these IL2dependent CTL which exhibit NK-like MHC-unrestricted cytotoxicity employ NKR structures in the recognition of tumor target cells. DBA/2.1 cells display both T lymphocyte and NK cell characteristics, i.e., they rearrange and express TcR genes; and are CD3+, Thyl.2'; but are CDS; CD8- and express the NK-associated markers NK2 and asialoGM1; and are highly cytotoxic for NK-sensitive YAC-1 tumor target cells (24). Upon completion of the immunization protocol, immune rat spleen cells were prepared and fused with SP2/0 myeloma cells. The resulting hybridomas were plated at a dilution calculated to yield one clone per well. Once growth was observed the clones were immediately subcloned at 0.3 cells/well. As shown in Table I, of the 12 subclones tested, culture supernatants from hybridoma S1C4 consistently showed significant (pc 0.01) inhibitory activity on DBA/2.1mediated cytolysis of YAC-1 target cells. The culture supernatants of hybridoma S1C4 were analyzed by ELISA and found to contain approximately 2 pg/ml IgM antibody. Given that the DBA/2.1 cells used for immunizations are not true NK cells but are best described as NK-like Cm, it was important to determine whether mAb S1C4 also inhibited the function of fresh splenic NK cells. The data shown in Table I1 indicate that endogenous NK activity of fresh DBA/2 (expt. 1) or CBA/J (expt. 2) spleen cells is partially inhibited by the presence of mAb SlC4. The inhibitory effect of mAb S1C4 on NK activity is dose-dependent, decreasing to negligible levels at mAb concentrations below 0.25 pg/ml (data not shown). In contrast to anti-LFA-1 mAb which was used as a positive control and had inhibitory effects on both NK and CTL activity, mAb S1C4 had no effect on killing mediated by alloantigen-specificCTL. The addition of an irrelevant control rat IgM to the assay had little or no effect on either NK or Cll function. Conjugate inhibition assays were performed to determine whether mAb S1C4 acts by interfering with effector-target cell conjugate formation. As shown in Table III, expt. 1, the presence of mAb S1C4 results in a 45% inhibition of conjugate formation between fresh spleen cells and YAC-1 target cells. The inhibitory effect is mediated at the level of the effector cell since pretreatment of spleen cells with mAb S1C4 followed by washing effectivelyinhibited conjugate formation to the same extent as when mAb S1C4 was present throughout the assay (expt. 2). Pretreatment of target cells had no effect on killing. Cellular distribution of the S1C4 anti=: Next we employed a cellular ELISA to examine the reactivity of mAb S1C4 with various murine lymphoid and non-lymphoid cell populations. This assay has previously been shown to be at least as sensitive as immunoperoxidase staining and fluorescence-activated cell sorter analysis (23). Representative data are shown in Table IV. As expected, mAb S1C4 was found to react stronglywith the immunizingDBA/2.1 NK-like CTL line and with asialoGM1' splenocytes, as well as with unselected DBA/2 spleen cells. Binding of mAb S1C4 to spleen cells of
HOSKIN A N D RODER
16
TABLE I Inhibitory Effect of S1C4 Hybridoma Supernatant on NK Activity of DBA/2.1 Cells
Immunol Invest Downloaded from informahealthcare.com by Stanford University on 12/08/14 For personal use only.
Culture supernatants added to assay"
% Cytotoxicity E/T 5:l
% Inhibitionb
22 27 19 16 21 26 22 23 20 22 25 25 23
SP2/0 control S1B9 SIC2 S1C4
SlFl S1G7 S1H7 S2A8 S2E12 S2H12 S3C10 S3G6 S4G2
0 14 27c 5 0 0 0 9 0 0 0 0
a 25,000 DBA/2.1 effector cells per microtiter well were pre-incubated for 30 min at 37°C with a 35 dilution of culture supernatants from SP2/0 myeloma cells or anti-DBA/2.1 hybridomas, followed by the addition of 5,000 "Cr-labeled YAC-1 target cells per well to give an E/T ratio of 51. % Inhibition = % cvtotoxicity DBA/2.1 cells + hvbridoma supernatant X 100 % cytotoxicity DBA/2.1 cells plus SP 2/0 supernatant 'P
