[Part 1: Biological Sciences] || Selective Blockade of Human Natural Killer Cells by a Monoclonal Antibody

Selective Blockade of Human Natural Killer Cells by a Monoclonal AntibodyAuthor(s): Walter NewmanSource: Proceedings of the National Academy of Sciences of the United States of America,Vol. 79, No. 12, [Part 1: Biological Sciences] (Jun. 15, 1982), pp. 3858-3862Published by: National Academy of SciencesStable URL: http://www.jstor.org/stable/12731 .

Accessed: 08/05/2014 18:53

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

National Academy of Sciences is collaborating with JSTOR to digitize, preserve and extend access toProceedings of the National Academy of Sciences of the United States of America.

http://www.jstor.org

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 18:53:25 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=nas

http://www.jstor.org/stable/12731?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Proc. Natl. Acad. Sci. USA Vol. 79, pp. 3858-3862, June 1982 Medical Sciences

Selective blockade of human natural killer cells by a monoclonal antibody

(receptors/target antigens/erythroid-myeloid cells)

WALTER NEWMAN

Program in Basic Immunology, Fred Hutchinson Cancer Research Center and Department of Microbiology and Immunology, University of Washington, Seattle, Washington 98104

Communicated by Eloise R. Giblett, March 15, 1982

ABSTRACT A murine monoclonal antibody, 13.1, which blocks human natural killer (NK) cell-mediated lysis, has been developed. Hybridoma 13.1 was derived by fusion of NS-1 cells with spleen cells from mice immunized with an enriched population of NK cells. Supernatants of growing hybridomas were screened for their ability to block NK cell-mediated lysis of K562 targets. Antibody 13.1 is an IgGl with a single light chain type and it does not fix complement. The 13.1 antigen is expressed on all peripheral blood mononuclear cells, with an antigen density approximately 1/30th that of HLA antigen heavy chain. Pretreat- ment and washing experiments revealed that inhibition of cytotoxicity occurred at the effector cell level only. Significant blocking was achieved with nanogram quantities of antibody and was not due to toxic effects on NK cells. Likewise, controls with other antibodies of the same subclass demonstrated that blocking was not a consequence of mere binding to NK cells. When a panel of 17 NK cell-susceptible targets was tested, the lysis of only 5 of these was blocked, namely K562, HL-60, KG-1, Daudi, and HEL, a human erythroleukemic cell line. The lysis of 12 human B cell and T cell line targets was not inhibited. In addition to the demon- stration that the 13.1 antigen is a crucial cell surface structure involved in NK lysis, a heterogeneity of target cell recognition has been revealed that argues for the proposition that individual NK cells have multiple recognitive capabilities.

Persuasive evidence culled from in vivo experiments demon- strates a role for natural killer (NK) cells in resisting the growth of tumors (1-3). Additional experiments with targets derived from a variety of embryonic or undifferentiated normal tissues suggests NK cells may also regulate host cell differentiation (4-6). Though not all cells are lysed by NK cells, the wide variety of tumor targets-tumors of lymphoid, myeloid, erythroid, embryonal, and fibroblast lineages, as well as xenogeneic tumors-raises the issue of whether there is any specificity to the binding and lytic phases. Unlabeled target inhibition and monolayer absorption experiments are consistent with a mul- tiplicity of NK target structures (7-9). However, Collins et al. have suggested that the susceptibility of fibroblasts to NK cell lysis may be determined primarily by the lack of an adequate target repair mechanism (2). A major unresolved issue, therefore, is the nature of the NK cell recognition process.

The use of antibodies, in the absence of complement, to block murine cytotoxic T cell function at either the effector (10-12) or target (13) cell level has provided information on both can- didate T cell receptors and target cell fine structures. This ap- proach has not yet been fully exploited for the examination of NK cell recognition or target structures.

An earlier report from this laboratory described the use of antibody 9.6, directed to the human T cell sheep erythrocyte

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertise- ment" in accordance with 18 U. S. C. ? 1734 solely to indicate this fact.

(E) receptor (14), to block NK cell function (15). Those results have been extended in the present report, which describes another monoclonal antibody 13.1, selected for its ability to block completely NK cell-mediated lysis at the effector cell level. Of interest here is the selective ability of antibody 13.1 to block lysis of some targets and not others.

MATERIALS AND METHODS Cell Preparation. Peripheral blood mononuclear cells (PBL)

were obtained from heparinized blood of normal volunteers by FicolVHypaque centrifugation. Cells were washed twice in Eagle's basal medium and resuspended in the same medium with 10% calf serum. A portion of such cells was frozen in 10% dimethyl sulfoxide/20% fetal calf serum (voVvol) in RPMI 1640 medium under liquid N2 for later use in 51Cr release assays.

51Cr Release Assay. NK and antibody-dependent cell-mediated cytotoxicity (ADCC) assays were performed as described (15). Briefly, effector PBL (10) were placed in V-bottom microtiter wells with 51Cr-labeled target cells (usually 103). Plates were spun gently and incubated at 37?C for 4 hr. One half of the supernatant (100 ,u) was harvested and its radioactivity was measured in a gamma counter. Results are expressed as % specific lysis based upon the formula [(experimental - spontaneous)/(freeze thaw - spontaneous)] x 100.

For screening of hybridoma supernatants, effector cells were obtained by thawing aliquots from a large pool of PBL obtained for this purpose by leukopheresis and frozen in 10% dimethyl sulfoxide under liquid N2. To optimize NK activity, PBL were first cultured overnight in RPMI 1640 medium/10% pooled human serum from nontransfused males; the medium contained poly(IC) (Sigma) at 50 Ag/ml for induction of interferon (IFN). Prior to assay, cells were washed twice and resuspended in Eagle's basal medium/ 10% calf serum. Recovery of cells ranged between 40% and 60%, with excellent viability and lytic activity.

Cell Lines. The cells used as targets were K562, Daudi, Raji, KG-1, HL-60, HPB-ALL, MOLT 4, CCRF-CEM, CCRF- HSB-2, RPMI-8402, Jurkat, Chang, Ramos, Nalm-6, and three B lymphoblastoid cell lines-GL, LD, and BC-derived from cells of normal individuals by transformation with Epstein-Barr virus. The KG-1 and HL-60 cell lines were obtained from M. Fukuda. The Daudi cell line was obtained from Barry Bloom, and the remaining lines were obtained from John A. Hansen. The Daudi line was tested to establish that it lacked cell surface HLA antigen and retained expression of human Ia antigen as

Abbreviations: NK, natural killer; ADCC, antibody-dependent cell- mediated cytotoxicity; PBL, peripheral blood mononuclear cells; HAT, hypoxanthine/aminopterin/thymidine; FACS, fluorescence-activated cell sorter; FITC, fluorescein isothiocyanate; IFN, interferon; ALL, acute lymphocytic leukemia; E, sheep erythrocyte; CTL, cytotoxic T lymphocyte(s).

3858



Medical Sciences: Newman Proc. Natl. Acad. Sci. USA 79 (1982) 3859

revealed by complement-mediated lysis in the presence of monoclonal antibodies W6/32 (16) and 7.2 (17), respectively. Human erythroleukemia cell line HEL, which is inducible for hemoglobin expression, was kindly made available by Paul Martin.

Immunizations and Fusion. The immunizing cells were pre- pared from human PBL enriched for NK activity by pretreatment with a combination of monoclonal antibodies 7.2 and 9.3 (17) plus complement. Simultaneous treatment with these antibodies plus complement results in lysis of 80-90% of PBL with no loss in NK activity (15). Dead cells were removed in two steps. Cells were first cultured overnight in the presence of DNase I (Sigma) at 1 ,tg/ml and then centrifuged on Ficoll/ Hypaque to remove cell debris. The resulting population (>95% viable) was washed twice in phosphate-buffered saline and 107 cells were injected intraperitoneally into 6-week-old BALB/c female mice. Animals were boosted twice at 10-day intervals. Three days after the last injection the spleens were excised, minced to single-cell suspensions, and fused with logarithmic-phase NS-1 cells as described by Kohler and Milstein (18). The ratio of spleen cells to NS-1 cells was 4:1. Fusion was accomplished with 40% (vol/vol) polyethylene glycol 1,500 in RPMI 1640 medium at 37?C by centrifugation and overlay of the cell pellet with polyethylene glycol. Fused cells were cultured with a feeder layer of BALB/c thymocytes (2 x 105 per well) in medium consisting of 100 ,uM hypoxanthine, 0.4 AM aminopterin, and 16 AM thymidine (HAT) and 10% fetal calf serum in RPMI 1640 medium. Supernatants were harvested.for screening on day 14. Thereafter cultures were refed with medium devoid of aminopterin (HT medium).

Hybridoma cells of interest were transferred to 16-mm Lin- bro wells in 2.0 ml of HT medium and after 4 days an aliquot of cells was frozen and a second portion was plated in microtiter plates at 10 cells per well (minicloning) on a feeder layer of thymocytes (5 x 10 per well). Two weeks later wells were again screened for blocking activity. Positive cultures were cloned in RPML 1640/10% fetal calf serum at 0.3 cells per well in a flat- bottom microtiter plate with 5 x 105 thymocytes per well. The cloning efficiency varied with the source of fetal calf serum and ranged between 10% and 25%.

Screening Assay. Blocking antibodies were detected by vir- tue of their ability to reduce the amount of 51Cr released in an NK assay using PBL effectors and K562 target cells. The effector-to-target ratio was 100:1. Effector cells were incubated with hybridoma supernatants for 1/2 hr at room temperature; target cells were then added. Plates were spun briefly and incubated for 4 hr at 37?C, supernatants were harvested, and radioactiv- ities were measured.

Reagents. Antibody 9.3 (17) recognizes a subpopulation (60-80%) of E rosetting cells and either is not expressed or is expressed at only low density on the majority of NK cells lytic for K562 (15). Antibody 7.2 (17) recognizes a framework region of the human Ia antigen and is undetectable on NK cells (15). Antibody 9.6 (14) is a mouse monoclonal antibody that blocks E rosetting of human T cells and blocks NK cell-mediated lysis of K562 (15). This antigen is easily detected on the great majority of NK cells (15). Antibodies 7.2, 9.3, and 9.6 were gen- erous gifts from John Hansen. Rabbit anti-Chang serum was a gift from Chris Henney. MOPC-21 myeloma protein, subclass IgGl/K, was purchased from Litton Bionetics (Kensington, MD). Anti-HLA monoclonal antibodies W6/32 (IgG2a) (16) and 34/28 (IgGi) (19) were purchased from Pel-Freez. Goat antisera to the subclasses of mouse immunoglobulin were purchased from Research Products International (Mt. Prospect, IL) and Litton Bionetics. Fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG was.purchased from Litton Bionetics.

Antibody Purification. BALB/c mice were primed by intraperitoneal injection of 1.0 ml of pristane (Aldrich) 6-10 days prior to intraperitoneal injection of 2 x 107 cloned hybridoma cells. Ten to 14 days later, ascites fluids were collected, spun to remove cells, and sterilized by filtration. Antibody 13.1 was purified in a single step by absorption and elution from staph- ylococcal protein A coupled to Sepharose (Sigma). Ascites fluid was applied to a column of protein A-Sepharose in 0.1 M Tris HCl, pH 8.1, at 4?C. The column was washed with buffer until the eluate registered an A280 of <0.05. Antibody was eluted with 0.1 M citrate/phosphate buffer at pH 5.5, neu- tralized immediately, and monitored for A280. Antibody con- centration was calculated on the basis of an A"m of 14.0.

Fluorescence Analysis. Cells (3 X 106) were incubated at 4?C with antibody 13.1 at 100-500 ,tg/ml for 1/2 hr, washed twice, and incubated with FITC-conjugated goat anti-mouse IgG for 1/2 hr at 4?C. Cells were washed twice, filtered to remove clumps, and analyzed on a fluorescence-activated cell sorter (FACS) II instrument. Each histogram presents results with 104

cells. The data are plotted in a logarithmic scale having a range of 104. Controls consisted of cells treated with MOPC-21 myeloma protein plus FITC-conjugated goat anti-mouse reagent. Glutaraldehyde-fixed chicken erythrocytes were used as a fluorescence standard. The relative fluorescence intensities were determined on the basis of the channel number of the geometric mean of a particular sample's fluorescence intensity. The channel number was compared for relative fluorescence intensity to glutaraldehyde-fixed chicken erythrocytes, arbitrarily set at 100, and.in this manner relative fluorescence intensities of different antibodies could be compared.

Two-Dimensional Gel Analysis. Protein A-Sepharose affin- ity-purified 13. 1 antibody (30 ,g) was eluted into sample buffer. For the first dimension of analysis, nonequilibrium pH gradient electrophoresis was performed with the gel system of O'Farrell (20). Gels were formed in glass tubes (2.5 X 130 mm) with a mixture of 9 M urea, 4% acrylamide, 0.11% bisacrylamide, 1% Nonidet P40, and 2.0% ampholine, pH 3.5-10 (Bio-Rad). Elec- trophoresis was performed at 500 V for 4 hr. After the run, gels were extruded into NaDodSO4 equilibration buffer containing 2-mercaptoethanol and incubated for 1/2 hr. Samples were then separated in the second dimension by NaDodSO4/polyacryl- amide gel electrophoresis using 10% acrylamide vertical slab gels with a 5% acrylamide stacking gel. The tube gel was fixed on top of the stacking gel by using a 1% agarose gel containing 2.3% NaDodSO4, 1% glycerol, 5% 2-mercaptoethanol, and 0.1% bromophenol blue in 62.5 mM Tris HCl, pH 6.8. Elec- trophoresis was performed at a constant voltage of 500 V, at 4?C, until the dye marker left the gel. The gel was stained with 0.125% Coomassie blue in methanol/acetic acid/water, 5:1:5 (vol/vol).

RESULTS Isolation of Antibody 13.1. Fig. 1 shows results from a portion

of the screening of fusion 13. Controls show that HAT medium had no effect on the NK assay compared to normal medium. In addition, antibody 9.6 as a positive control exhibited approximately 50% blocking, and MOPC-21 had no effect at 0.1 mg/ ml. As can be seen from the results obtained in column 6, well E6 exhibited significant reduction in the amount of 51Cr released from K562 target cells. Hybridoma cells from well E6 have been cloned seven times by limiting dilution on a feeder layer of thymocytes.

Characterization of Antibody 13.1. Antibody 13.1 has been examined in gel diffusion with subclass-specific antisera and found to react specifically with anti-mouse IgGl. Consistent with this subclass designation is the observation that 13.1 can



3860 Medical Sciences: Newman Proc. Natl. Acad. Sci. USA 79 (1982)

Controls MOPC-21

Medium Spontaneous

HAT Antibody 9.6

Total Fusion 13, column 6

Row A Row B Row C Row D Row E Row F Row G Row H

0 2 4 6 8 10

cpm x 10-2

FIG. 1. Screening assay for detection of blocking antibodies. At top are shown controls performed to detect possible nonspecific inhibition of lysis by HAT medium or irrelevant antibody (MOPC-21). Specific inhibition is detected in culture supernatants of clone 9.6 at 5 .Ag/ml. Below are the results from one row of fusion 13, demonstrating inhibition of 51Cr release by the supernatant of well E6.

be eluted from protein A-Sepharose at pH 5.5 (21). The antibody was further analyzed by two-dimensional gels. As shown in Fig. 2, 13.1 possesses single heavy and light chains.

Expression of 13.1 Antigen on PBL. Fig. 3 shows the relative fluorescence intensities of staining of PBL with 13.1 compared to two other monoclonal antibodies: 9.6, an anti-E receptor, and 34/28 (19), an IgGl antibody specific for HLA. All antibodies were tested at saturating concentrations. As can be seen, all PBL stained with 13.1 with a weak fluorescence intensity. An- tibody 34/28 stained with approximately a 30-fold greater intensity than 13.1 and antibody 9.6 stained with approximately a 6-fold greater intensity than 13.1. Direct examination of 13.1- stained PBL in the fluorescence microscope revealed that less than one in 500 cells lacked 13.1 antigen expression.

Effect of Antibody 13.1 on NK Cell-Mediated Lysis. Purified antibody 13.1 was tested for ability to block lysis of K562 targets at a range of antibody concentrations. As shown in Fig. 4, antibody at 1 ng/ml reduced lysis of K562 by approximately 50%, with blocking still detectable at 0.1-0.2 ng/ml. Incubation of cells with antibody 34/28 at 12.5 Ag/ml had no effect on NK

^ - TEF

I

FIG. 2. Two-dimensional gel analysisof-monoclonal antibody 13. 1.

|~~~~~~~~~~~~~-L _ 4r|

can L, lih chain.

I '

1 10 100 1,000

Fluorescence intensity

FIG. 3. FACS profiles of peripheral blood lymphocytes stained with MOPC-21 plus FITC-labeled goat anti-mouse IgG (negative control) (... ), 13.1 (--- ), 9.6 anti-E receptor (-), or 34/28 IgGl anti- HLA (.-.). Fluorescence intensity is plotted on a logarithmic scale in arbitrary units, with 100 representing the mean fluorescence intensity of glutaraldehyde-fixed chicken erythrocytes.

cell lysis. Identical results have been obtained with five additional donors. Hence, blocking was achieved at low concentrations of antibody 13.1, is not due to excessive coating of effector cells with antibody, and is not a general property of murine IgGl antibodies that bind to human PBL.

In order to address whether the blocking observed was accomplished at the effector or target cell level, PBL or K562 targets were preincubated with antibody, washed three times, then tested. Results from three experiments are shown in Table 1. Clearly all blocking activity can be accounted for by antibody binding to the 13.1 antigen on effector cells.

Selectivity of 13.1-Mediated Blocking. The ability of 13.1 antibody to block lysis was tested on a panel of 17 NK cell-susceptible targets. Effector cells from 15 unrelated individuals were tested in 20 separate experiments. The results shown in Table 2 for each cell line are representative experiments. Each target was tested with at least five different donors.

Effector cells were blocked in their lysis of the erythroleukemia target cells K562 and HEL, the myelocytic leukemia cells KG-1, and the promyelocytic leukemia cells HL-60 (Table 2A). No blocking of lysis of six different T lymphoma cell lines was observed, nor was there any effect on ADCC with antibody- coated Chang liver cell targets.

Table 2B presents results from the NK cell-mediated lysis of B cell line targets. In order to achieve sufficient lysis, the effector cells were preincubated for 18 hr in the presence of the

70 -

60

50

< 40

, 30

8 20

10

0 103 10 1O-1 10-3 io-5

Antibody 13.1, ng/ml

FIG. 4. Dose-response curve for blocking by antibody 13.1. Each point represents the percent specific lysis by 2 x 10r PBL of 2 x 103

Cr-labeled K562 targets, mean ? SEM, in the presence of various concentrations of antibody 13.1. The broken line represents the percent specific lysis in the presence of monoclonal anti-HLA, 34/28, at 1:2.5 ,u&g/ml.



Medical Sciences: Newman Proc. Natl. Acad. Sci. USA 79 (1982) 3861

Table 1. Effect of pretreatment of effector cells or target cells with antibody 13.1 on NK cell-mediated lysis

% specific lysis Effectors + Targets, +

Exp. Target MOPC-21 13.1 13.1 13.1 1 K562 71 11 13 63 2 K562 72 4 7 68 3 K562 68 16 20 63

Effector PBL (3 x 106) or 51Cr-labeled targets (3 x 106) were incubated with 10 ,g of 13.1 antibody or 10 Ag of MOPC-21 in 1.0 ml of Eagle's basal medium/10% calf serum. After 1/2 hr at room temperature, cells were washed three times, counted, and added to microtiter plates at an effector-to-target cell ratio of 50:1. Results are compared to direct addition of antibody 13.1 to assay wells at 10 Ag/ml.

IFN inducer poly(IFC) to augment NK activity. Despite aug- mentation of lysis, blocking of K562 killing, was still observed. Of the B cell lines tested, only lysis of the Daudi cell line was reproducibly inhibited. Lysis of Raji, another Burkitt lymphoma-derived line, was not affected, nor was the lysis of the Epstein-Barr virus-negative Burkitt-derived cell line Ramos. No blocking of the lysis of Nalm-6-, a pre-B leukemic line, or of three Epstein-Barr virus-transformed B cell lines from normal individuals was observed. The results presented in Table 2 were reproducible over a range of antibody concentrations

Table 2. Effect of antibody 13.1 on NK cell-mediated lysis

% specific lysis % inhi- Target Lineage Medium 13.1 bition

A. Lysis.of various target cells K562 Erythroleukemiat 69 15 79 HEL Erythroleukemia 67 16 76 KG-1 Myelocytic 22 8 64

leukemia HL60 Promyelocytic 34 18 47

leukemia

HPB-ALL ALL-T 97 96 1 MOLT 4 ALL-T 42 43 -2 CCRF-CEM ALL-T 53 53 0 CCRF-HSB-2 ALL-T 33 39 -18 RPMI-8402 ALL-T 20 20 0 Jurkat ALL-T 48 48 0

Chang + anti- Chang serum ADCC 45 41 9

B. IFN-augmented;lysis of B cell targets K562 Erythroleukemia 79 36 54 Daudi Burkitt lymphoma 71 42 41 Raji Burkitt lymphoma 23 26 -13 Ramos Burkitt lymphoma 52 49 6 Nalm-6 Pre-B leukemia 55 63 -15 GL B lymphoblastoid 34 34 0

cell line LD B lymphoblastoid 36 36 0

cell line BC B lymphoblastoid 36 38 -6

cell line

InA, 2 x 105 effector cells were treated with' antibody 13.1 at 10 Ag/ ml in microtiter plates, followed by addition of 2 x 103 51Cr-labeled targets. ALL, acute lymphocytic leukemia. In B, effector cells were incubated overnight in RPMI 1640 medium/10% pooled human serum containing poly(T1C) at 50 ,ag/ml to augment NK lysis. Cells -were washed twice prior to assay and- preincubated for 30 min with antibody 13.1 at 10 ,ug/ml, followed by addition of 51Cr-labeled targets.

30

u'20

10-

Unlabeled Jurkat Unlabeled K562

0 2.5 5 10 20 2.5 5 10 20

- Ratio of unlabeled to labeled targets

FIG. 5. Effector PBL were incubated in the presence of MPOC-21 at 10 Ag/ml (o) or 13.1 at 10 ,g/ml (s} for 1/2 hr at 22?C. Thereafter unlabeled Jurkat (Left) or K562 (Right) targets were added to all wells- at various doses, followed by addition of 2 x 103 51Cr-labeled Jurkat targets. Lysis in the absence of competitors was 25%.

between 1 ,ug/ml and 1 ng/ml and at effector-to-target ratios of 50:1 and 25:1 (data not shown).

Analysis by Inhibition of Lysis by Unlabeled Targets. To address whether there was an effect on the NK-K562 interactions, advantage was taken of two. observations: first, that lysis of the T lymphoma line Jurkat was unaffected by antibody 13.1; and, second, that K562 is an effective competitor of the lysis of Jurkat cells. As shown in Fig. 5, NK cell-mediated lysis of Jurkat cells in the presence of control antibody MOPC-21 was blocked equally well by unlabeled Jurkat cell or unlabeled K562 competitors, suggesting competition for a common pool of NK effector cells. Fig. 5 also shows the same experiment except that the effector cells were preincubated with antibody 1'3.1' and washed to remove unbound antibody. As can be seen in this experiment (one of five with identical results), Jurkat cells con- tinued to be effective competitors of'the lysis of 51Cr-labeled Jurkat cells, but unlabeled K562 cells competed poorly, suggesting that antibody 13.1 acts to prevent the ability of K562 cells to compete with Jurkat cells for a population of NK cells lytic for both K562 and Jurkat targets.

DISCUSSION This report describes a murine monoclonal antibody that, in the absence of complement, blocks human NK cell-mediated lysis. The antibody is;an IgGI, has a single species of light chain, and does not fix complement., Blocking. is accomplished by nanogram quantities of antibody, which argues against "coating" of effector cells as a means of inhibition. Mere binding of NK cells by murine IgGl antibody is not sufficient to cause inhibition. This was demonstrated with antibody 34/28, an IgGl specific for a monomorphic determinant of HLA (19). Antibody 34/28 failed to block NK cell lysis despite the fact that 30-fold more 34/28 than 13.1 antibody binds to PBL. In a previous publication it was also shown that. two other monoclonal reagents, W6/32 anti-HLA of the IgG2a subclass (16) and OKM1 (22), both of which bind to NK cells, do not block NK cell lysis in the absence of complement. Abo and Balch (23) have described a monoclonal antibody specific for human NK and ADCC cells, HNK-1, which in the absence of complement does not block NK cell lysis. Hence, antibody binding to NK cells is insuffi- cient to block lysis. Moreover, blocking is not due to toxic effects on NK' cells, because cytolytic activity is recoverable and antibody 13.1 does not block lysis of cytotoxic T lymphocytes (CTL) (unpublished results). Pretreatment of effector or target cells with antibody 13.1, followed by washing, revealed that blocking was achieved at the effector cell level.



3862 Medical Sciences: Newman Proc. Natl. Acad. Sci. USA 79 (1982)

Of special interest is the observation that antibody 13.1 is selective in its inhibitory ability. Only 5 of 17 targets tested were inhibited from lysis when a 1,000-fold excess of the amount of antibody required to achieve 50% inhibition of K562 lysis was used. Likewise, no inhibition of ADCC against Chang cells was observed. Those targets whose lysis was blocked were primarily of the myeloid and erythroid lineages. The one exception was the Daudi cell line. Lysis of other B cell targets, including two Burkitt lymphoma-derived B lines, was unaffected, as was the lysis of six different T cell lines derived from leukemia patients. Further studies will be necessary to ascertain whether there are any cell surface molecules uniquely expressed (or lacking) on those cell lines whose lysis is inhibited.

To analyze the mechanism by which lysis is blocked, advantage was taken of two observations: first, that K562 cells effec- tively compete with 5"Cr-labeled Jurkat targets for NK cells; and, second, that antibody 13.1 'blocks lysis bf K562 and does not block lysis of Jurkat targets. Hence, it was feasible to ask whether NK lysis of Jurkat cells was inhibited as well by K562 in the presence as in the absence of antibody 13.1. Whereas K562 targets were rendered inefficient competitors by pretreatment of NK cells with antibody 13.1, no effect on the competitive ability of Jurkat cells was observed. Moreover, the ability of K562 to compete with' Jurkat cells for NK cells in the absence of antibody 13.1 indicates that these targets compete for the same pool of effector NK cells.

This set of observations is best explained by postulating that NK cells possess multiple recognition structures. If lysis of these two targets was achieved by distinct NK cells with different receptors, then K562 would not be an efficient competitor of the lysis of Jurkat cells. Furthermore, it seems unlikely that NK cells recognize K562 and Jurkat targets via a common structure, given the selective inhibitory effects of antibody 13.1.

A model of multivalent NK cells is consistent with the observation that many different kinds of unlabeled targets can compete with a single 51Cr-labeled target (7, 24) and that NK specificity does not segregate upon cloning (25). The multivalent model does not require that all NK cells possess the same set of receptors; subsets of NK cells may differ in the distribution of receptor expression. This might explain the occasional ability to adsorb NK effectors on monolayers of targets in a specific manner (26). However, validation of the multireceptor hypothesis will have to await experiments with cloned human NK cells.

The 13.1 antigen is similar to the murine Ly-5 antigen in its expression on all lymphoid cells and in its ability to block NK cell but not CTL-mediated lysis (27). A murine antigen, H-ll, which is distinct from Ly-5, has recently been described; antibody binding to this determinant also selectively blocks NK and not CTL activity (28). We have previously demonstrated that antibody 9.6 anti-E receptor (14) blocks human NK as well as CTL-mediated lysis (15). The 13.1 antigen is distinct from the 9.6 antigen as judged on the basis of competitive binding assays, tissue distribution, and failure of the 13.1 antibody to block E-rosetting (unpublished results). Hence, there are at least two distinct cell surface structures on both human (9.6 and 13.1) and murine (Ly-5 and H-lI)'NK cells that are intimately involved in the recognition or lytic phases of NK cytolysis. In this context, the observation that the 13.1 antigen is expressed

on all lymphoid cells argues against a receptor function for the 13.1 molecule. More likely it is proximate to and interferes with some protein with receptor capability. Alternatively, the 13.1 antibody may be crossreactive with an NK receptor structure. Although the molecular weight of the 13.1 antigen remains to be established, it is sensitive to trypsin and Pronase, and its expression on the cell surface is resistant to tunicamycin treatment (unpublished results). These and other monoclonal antibodies should,help to analyze the complexities of NK-target cell interactions.

I acknowledge the expert technical assistance of Ms. G. Shu. I thank Drs. Colin Brooks and Paul Martin for helpful discussions. This work was supported by U.S. Department of Health and Human Services Grant AI-16496 and National Science-Foundation Grant PCM-8008749.

1. Haller, O., Hansson, M., Kiessling, R. & Wigzell, H. (1977) Na- ture (London) 270, 609-611.

2. Collins, J. L., Patekand, P. Q. & Cohn, M. (1981)J. Exp. Med. 153, 89-106.

3. Kasai, M., LeClure, J. C., McKay-Bourdreau, L., Shen, F. W. & Cantor, H. (1979) J1. Exp. Med. 149, 1260-1264.

4. Cudkowicz, G. & Hochman, P. S. (1979) Immunol. Rev. 44, 13-41.

5. Stern, P., Gidlund, M., Orn, A. & Wigzell, H. (1980) Nature (London) 285, 341-342.

6. Hansson, M., Kiessling, R. & Andersson, B. (1981) Eur. J. Im- munol. 11, 8-17.

7. Welsh, R. M. & Zinkernagel, R. M. (1977) Nature (London) 268, 646-648.

8. Herberman, R. B., Nunn, M. E. & Lavrin, D. H. (1975) Int. J. Cancer 16, 216-229.

9. Clark, E. A. & Sturge, J. C. (1981)J. Immunol. 126, 969-974. 10. Shinohara, H. & Sachs; D. H. (1979) J. Exp. Med. 150, 432-444. 11. Hollander, N., Pillemer, E. & Weissman, I. L. (1980) J. Exp.

Med. 152, 674-687. 12. Davignon, D., Martz, E., Reynolds, T., Kurzenger, K. & Spring-

er, T. (1981) J. Immunol. 127, 590-595. 13. Lindahl, K. F. & Lemke, H. (1979) Eur.J. Immunol. 9, 526-536. 14. Kamoun, M., Martin, P. J., Hansen, J. A., Brown, M. A., Sia-

dak, A. W. & Nowinski, R. C. (1981) J. Exp. Med. 153, 207-212. 15. Fast, L. D., Hansen, J. A. & Newman, W. (1981) J. Immunol.

127, 448-452. 16. Barnstable, C. J., Bodmer, W. F., Brown, G., Galfre, G., Mil-

stein, C., Williams, A. F. & Ziegler, A. (1978) Cell 14, 9-20. 17. Hansen, J. A., Martin, P. J. & Nowinski, R. C. (1980) Immuno-

genetics 10, 247-260. 18. KUhler, G. & Milstein, C. (1975) Nature (London) 256, 495-497. 19. Truco, M. M., Garotta, G., Stocker, J. W. & Ceppellini, R.

(1979).Immunol. Rev. 47, 219-252. 20. O'Farrell, P..Z., Goodman, H. M. & O'Farrell, P. H. (1977) Cell

12, 1133-1141. 21. Ey, P. L., Prowse, S. J. & Jenkin, C. R. (1978) Immunochemistry

15, 429-436. 22. Breard, J., Reinherz, E. L., Kung, P. C., Goldstein, G. &

Schlossman, S. F. (1980)J. Immunol. 124, 1943-1948. 23. Abo, T. & Balch, C. M. (1981)J. Immunol. 127, 1024-1029. 24. Herberman, R. B., Nunn, M. E. & Lavin, D. H. (1975) Int. J.

Cancer 16, 216-229. 25. Dennert, G., Yogeeswaran, G. & Yamagata, S. (1981) J. Exp.

Med. 153, 545-556. 26. Phillips, J. H., Ortaldo, J. R. & Herberman, R. B. (1980) J. Im-

munol. 125, 2322-2327. 27. Kasai, M., Leclerc, J. C., Shen, F. W. & Cantor, H. (1979) Im-

munogenetics (N.Y.) 8, 153-159. 28. Seaman, W. E., Talal, N., Herzenberg, L. A., Herzenberg, L.

A. & Ledbetter, J. A. (1981) J. Immunol. 127, 982-986.



Documents

[Part 1: Biological Sciences] || Selective Blockade of Human Natural Killer Cells by a Monoclonal Antibody