l 3:»: ‘ ‘ ‘ ‘-- ‘ ‘3‘ ”4 I 3 f2" «4 4111' w“ 31.. 11' H ,v-ndfi ~ - . Y‘IWW1KIW1 , 1 “my“ a: ”1:1“? 1‘ “11‘ Mg}. in; ”"1111" 91“ I ‘ L 9 :1 , \ .. ‘ 'JL’ "t’:‘r ‘ 719;" 131’, @131 1.11:” w: NEE 1‘13 ~ “..A' ‘1.” '11::5:1"‘:1 EI-‘Efi‘ li’f’”" $311131: . fr! $15. .‘1, 1 “ . " ’é’f“; ” -’;' figufimgiil’é. I . an l-i 4’1. 4‘ ‘ *‘Qi, u . k "4‘ aft-J- _. 3 Wm. ; 3:1... >- ‘-’ 7’ ‘a .p s " v95: 1- 14' 1‘: Ln 1 511233113 fin ‘ ~ 1L} 5"“ ’ 1L 111.71%" 11151 1 1.1% 1 12111111131131 ~ 11111513 111‘ ' 1’31 :91: 11111 1 ' 13’1”” ’ :11’1’11‘155‘151’ 1 ’1 11 c n,» 1; '133 (1’1“ 1;";1’111'11f‘, (572.111 1 1‘1 “WP? 1111,3111. 1’1‘111’ ‘11“ ““1 1111 ‘ “Tam . 1 ’ {3’15 11‘1““ ”3%" ’1? 113-11“. 55“ '1?» 1;, : fl?" W’fié’” Li :3. 3 1.5%; v. 4‘. t 1 ‘3 1 . 111’“ 42:91? " I ‘1‘ 1311“”: ‘ ‘11“1’1 .. “1,3“ :11 ‘ I111. J’l"”’V' (1’11 1311.}:3‘16‘” ’1’! 11'?! . 1%" 3’1”” ’1’”! ”11": " a . I 1 . 1 $111...ng 111-, (1,1)“ 171.? 1112.11 1’ | A 1”“: 1"11‘10‘ ’ H11111’1’!”9”” In 4 .11 1"‘1'1’11‘2” ’1 ":22; ‘5’: \ f‘ ‘1 Mr? Y:""‘=-V: \‘ '. '1 111.” 1””: {THESIS This is to certify that the i dissertation entitled SURFACE Ly—5 GLYCOPROTEIN IN MURINE NATURAL KILLER (NK) CELL DEVELOPMENT, TARGET BINDING AND CYTOTOXICITY: FUNCTIONAL STUDIES AND PRELIMINARY BIOCHEMICAL CHARACTERIZATION presented by Michael Harry Zaroukian has been accepted towards fulfillment of the requirements for __BhIJL_ degree in _Immunology_ October 31, 1984 Date M5u.'.m4rnn...m..‘ ' 'r '0”, 'J' ' ‘ 042771 MSU LIBRARIES _c—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. SURFACE Ly-S GLYCOPROTEIN IN MURINE NATURAL KILLER (NK) CELL DEVELOPMENT, TARGET BINDING AND CYTOTOXICITY: FUNCTIONAL STUDIES AND PRELIMINARY BIOCHEMICAL CHARACTERIZATION BY Michael Harry Zaroukian A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1984 ABSTRACT SURFACE Ly-S GLYCOPROTEIN IN MURINE NATURAL KILLER (NK) CELL DEVELOPMENT, TARGET BINDING AND CYTOTOXICITY: FUNCTIONAL STUDIES AND PRELIMINARY BIOCHEMICAL CHARACTERIZATION BY Michael Harry Zaroukian The role of surface Ly-S glycoprotein expression in the binding and lysis of susceptible tumor targets by natural killer cells was studied in murine splenocytes reacted with anti-Ly-S serum. Conjugate assays demonstrated that abrogation of tumor cell lysis by anti-Ly-S serum involved the inhibition of NK cell binding to susceptible YAC-l targets. Treatment of cell sorter purified Ly-S- cells with interferon converted 22% of the cells to a Ly-5+ phenotype and induced NK activity against YAC-l targets. Sorted Ly-S— cells cultured three weeks in interleukin and interferon containing medium (CM) acquired substantial surface Ly-S with concomitant high levels of NK activity that remained susceptible to inhibition by anti-Ly-S serum. To gain further insight into the physical properties of Ly-S glycoprotein as a candidate surface molecule for NK cell interaction with target cells, Michael Harry Zaroukian preliminary biochemical characterization of Ly-S on the plasma membrane of murine NK cell-enriched splenocytes was performed. Radioiodination of surface proteins followed by immunoprecipitation of Ly-S and polyacrylamide gel electrophoresis (SDS-PAGE) revealed a relative molecular mass (Mr) of 205,000 which was indistinguishable from that of macrophage-associated Ly-S. While trypsinization of viable cells did not diminish the amount of immunofluorescently detectable surface Ly—S, SDS-PAGE analysis of the radioiodinated glycoprotein revealed that it had been cleaved to a peptide with an approximate Mr of 125,000 which resisted further trypsinization. The peptide remained closely associated with the plasma membrane, did not exist as a disulfide—linked heterodimer and was released only after detergent solubilization of membrane proteins. The Mr of the immunoprecipitable fragments generated from trypsinized viable cells and NP-40 solubilized membrane proteins were identical, suggesting that the fragment lacked the intracytOplasmic peptide sequence of the intact Ly-S glycoprotein. The data presented suggest that surface Ly-S glyc0protein expression is important for binding of NK cells to YAC-l targets, that Ly-S- precursors of NK cells are present in murine splenic tissues and can be induced by CM to become highly active NK cells with increased surface Ly-S expression. The persistent susceptibility of these cells to inhibition of cytotoxic activity by anti-Ly-S serum and Michael Harry Zaroukian biochemical characteristics consistent with those of a putative NK cell receptor provide additional evidence of an important role for the Ly-S glyc0protein in the NK cell cytolytic mechanism. ACKNOWLEDGEMENTS I wish to express‘ my sincere appreciation to my academic mentor, Dr. Harold C. Miller, for valuable advice, support and assistance in my development as a research investigator, without which entrance into and completion of this degree program would not have been possible. The personal interest, guidance and encouragement of Dr. Walter J. Esselman, along with his thoughtful discussions and intellectual curiosity are gratefully acknowledged and shall serve as a standard of excellence in scientific method and the education of young scientists. I am grateful to the other members of my guidance committee, Dr. Maria Patterson, for her continued involvement in my academic development despite the rigorous demands of medical school training, and Dr. Pamela Fraker, who was instrumental in developing my ability to critically evaluate the scientific literature and generously agreed on short notice to serve on my guidance committee late in the course of the doctoral program. The willingness of Dr. Robert Brubaker to participate in my oral and written comprehensive ii examinations as well as serving as the Graduate Committee representative is appreciated. I also wish to thank Drs. Ruth Hoppe, Patrick Alguire and Raymond Murray of the Department of Internal Medicine, College of Human Medicine, for their willingness to participate in this combined educational process, for helping to advance my clinical and research skills as a physician while I received formal scientific training and for their moral support, flexibility and steadfast advocacy of my development as a clinical investigator and academic physician. Finally, to my wife Barbara, for her inestimable importance to my life and work, shall go my unending gratitiude. iii TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O O O O O O O O O O O V LIST OF FIGURES O O O O O O O O O O O O O O O O O O 0 .Vi LIST OF ABBREVIATIONS . . . . . . . . . . . . . . viii INTRODUCTION 0 O O O O O O O O O O O O O O O O O O O O O 1 LITERATURE REVIEW 0 O O O O O O O O O O O O O O O O O O 6 LITERATURE CITED. . . . . . . . . . . . . . . . . . . .33 ARTICLE 1 - SURFACE Ly-S GLYCOPROTEIN IN MURINE NATURAL KILLER (NK) CELL DEVELOPMENT, TARGET BINDING AND CYTOTOXICITY . . . . . .46 ARTICLE 2 — PRELIMINARY BIOCHEMICAL CHARACTERIZATION OF SURFACE Ly-S GLYCOPROTEIN ON MURINE NATURAL KILLER CELLS. . . . . . . . . . . .78 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . 116 iv LIST OF TABLES Table Page ARTICLE 1 I Effect of Anti-Ly-S Serum on NK Cell-Mediated Cytotoxicity against YAC-l Target Cells . . . 56 II Inhibitory Effect of Anti-Ly-S Serum on Target cell Binding by NK cells. 0 O O O O O O O O O 58 III Effect of Interferon Treatment on NK Activity of Sorted Ly-5+ and Ly-S- C57BL/6 splenocytes O O O O O O O O O O O O O O O O 0 64 IV Increased NK Activity of Sorted C57BL/6 Spleen Cells Cultured 3 Weeks in vitro in Conditioned Medium and Inhibition by Anti-Ly-S mAb. . . . 67 ARTICLE 2 I Enrichment of Murine NK Cell Activity by Percoll Discontinuous Density Gradient Centrifugation. . . . . . . . . . . . . . . . 91 Figure ARTICLE 1 1 ARTICLE 2 1 LIST OF FIGURES Fluorescence histograms of unfractionated (UF) and nylon wool nonadherent (NWNA), C57BL/6J splenocytes labeled with biotinylated anti-mouse IgG and fluoresceinated avidin . . . . . . . . . . . Fluorescence histograms of viable NWNA, anti-Thy 1 + GPS treated C57BL/6J splenocytes treated with anti-Ly-S mAb + biotin/avidin-FITC . . . . . . . . . . . . . Two-dimensional cytograms (narrow angle scatter versus right angle fluorescence) of sorted Ly-5- C57BL/6J splenocytes before and after 3 hour incubation with 5000 units murine interferon (ac +18) or medium control. Fluorescence histograms of sorted Ly-S- and Ly-5+ C57BL/6J splenocytes after 3 week culture in conditioned medium (CM) . Fluorescence histograms of control and trypsinized NK cell-enriched B6 splenocyte suspensions labeled with murine monoclonal anti-Ly-S, biotinylated anti-mouse IgG and fluoresceinated avidin . . . . . . . . . . . Autoradiographic analysis of SDS-poly- acrylamide gel electrophoresis of Ly-S immunoprecipitates from lactoperoxidase- catalyzed surface radioiodination of unfractionated and NK cell-enriched splenocyte suspensions . . . . . . . . . . . Effect of trypsinization of surface radio- iodinated 60% Percoll-fractionated B6 splenocytes . . . . . . . . . . . . . . . . vi Page . 59 .63 o66 ~93 .95 .97 Inability of simple sugars to release Ly- fragment from trypsinized cells. . . . . SDS-polyacrylamide gel autoradiograph of Ly-S immunoprecipitates from trypsinized 60% Percoll-fractionated splenocytes and trypsin supernatants . . . . . . . . . . Evidence that the 125 KD Ly-S fragment does not exist as a disulfide-linked heterodimer. . . . . . . . . . . . . . . Apparent molecular weight identity of the 125 KD Ly-S fragment generated by trypsinization of intact cells and NP-40 solubilized cell surface proteins. . . . vii 5 . 98 .100 .101 .103 LIST OF ABBREVIATIONS ADCC antibody dependent cellular cytotoxicity AK activated killer, anomalous killer avidin-FITC fluoresceinated avidin B6 C57BL/6J biotin/avidin—FITC sequential reaction with biotinyl- ated anti-mouse IgG and fluoresceinated avidin BSA bovine serum albumin C complement °C degrees Centigrade CM conditioned medium ConA concanavalin A cpm counts per minute CT chymotrypsin CTL cytotoxic T lymphocytes EDTA ethylenediaminetetraacetic acid EGTA ethylene glycol tetraacetic acid FBS fetal bovine serum FcR receptors for the Fc portion of Ig FCKR receptors for the Fc portion of IgG GPS guinea pig serum viii HEPES IFN 19 IL-2 IL-3 KCIL KD LDCC LGL LPO MHC MLC MLC-AK NC NK NKCF NP-4O NWNA PBL N-2-hydroxyethylpiperazine—N'-2- ethane-sulfonic acid immune response region associated interferon immunoglobulin interleukin-2 interleukin-3 killer killer cell independent lysis kilodaltons lectin dependent cellular cytotoxicity large granular lymphocytes lactoperoxidase molar monoclonal antibody 2—mercaptoethanol major histocompatability complex mixed lymphocyte culture mixed lymphocyte culture-activated killer relative molecular mass natural cytotoxic natural killer natural killer cytotoxic factor Nonidet-P40 nylon wool nonadherent papain peripheral blood lymphocytes ix PBS PHA PMSF SAC SDS SDS-PAGE 519 T TBC TCGF TEMED UF WGA phosphate buffered saline phytohemagglutinin phenylmethylsulfonylfluoride Staphlococcus aureus cells sodium dodecyl sulfate sodium dodecyl sulfate-polyacryl- amide gel electrOphoresis surface immunoglobulin trypsin tumor binding cells T cell growth factor, interleukin-2 N,N,N',N'-tetramethylethlyene— diamine unfractionated wheat germ agglutinin INTRODUCT ION The experiments described in this dissertation have been devised to identify and characterize developmental, biochemical and functional aspects of a relatively recently discovered cytotoxic effector cell, the natural killer (NK) cell. First identified in the early 1970's (Rosenberg, et a1, 1972; Kiessling, et al, 1975a), NK cells have been implicated in various aspects of host defense mechanisms against cancer (Herberman, 1981), microbial infections (Hatcher and Kuhn, 1982; Murphy and McDaniel, 1982) and in graft rejection (Kiessling, et a1, 1977). DeSpite the considerable efforts of numerous investigators exploring the nature and function of NK cells, relatively little is known about the progenitors of NK cells, the cell lineage to which NK cells belong, mechanisms leading to their functional maturation and cell surface molecules that are important in one or more steps of the cytolytic reaction sequence. The approach taken in the experiments described in the first article was to identify, on the basis of inhibition by specific antibody, a candidate surface molecule on natural killer cells whose expression is important to NK cell-mediated cytotoxicity against 1 malignant cells. A molecule identified in this manner and studied for its role in NK cell development and function was the Ly-S glycoprotein (also called T200 glyCOprotein or leukocyte common antigen). Ly-S comprises a molecularly heterogeneous group of glycoproteins found on cells of lymphoid, myeloid and monocytoid origin (Scheid and Triglia, 1979). Once identified as a protein that may participate in the NK cell-mediated lysis of tumor cells, Ly-S was further examined to determine whether it mediates any of the steps of the cytolytic mechanism, namely, target cell binding, programming for lysis or killer cell—independent lysis (Hiserodt, et a1, 1982). The results from conjugate assays indicated that Ly-S glycoprotein acts as a receptor on NK cells that mediates binding to the YAC—l lymphoma target cell. In order to identify precursors of NK cells and those processes leading to their functional maturation to cytotoxic effector cells, it was postulated that the surface Ly-S glycoprotein which appeared to be important for NK activity might be lacking from the surface of pre-NK cells. This hypothesis was tested by fluorescence—activated cell sorting of murine splenocyte suspensions into highly purified Ly-5+ and Ly-S— populations. Sorted cells were tested for NK activity before and after treatment with agents that have been shown to augment NK activity, such as interferon (IFN) and interleukin-2 (IL-2; Brooks, 1983). Observed differences in NK activity were correlated with changes in surface Ly-S glycoprotein expression occuring under the same conditions. Additional aliquots of sorted cells were placed into three week in yitrg culture in medium containing IFN and IL-2 and compared for Ly-S expression and NK activity. The results of these experiments indicated that Ly-5- pre-NK cells are present in murine splenic tissues and that IFN- and IL-2—containing preparations induce conversion of NK-inactive Ly-S- pre-NK cells to potently cytolytic NK cells bearing very high surface levels of Ly—5 glycoprotein. Even at a low effector:target ratio, the cytotoxic activity of in yitrg cultured NK cells greatly exceeded that of freshly isolated NK cells. This activity was significantly inhibited by anti-Ly-5 serum, further supporting a role for the Ly-S glycoprotein in the cytolytic process. Having established evidence for the importance of Ly-5 in murine NK cell development, target cell binding and cytotoxicity, it became of interest to biochemically characterize the glycoprotein on NK cells. The molecular weight heterogeneity of Ly-5 observed between purified and cloned populations of T cells, B cells and monocytes (Trowbridge, et al, 1975; Tung, et a1, 1981), raised the question of whether the molecular weight form of Ly-S on NK cells is similar to that found on other lymphoid or myeloid populations or if instead it comprises a unique form of the glycoprotein. Experiments described in the second article were designed to identify the apparent molecular weight of Ly-5 glycoprotein obtained from NK cell-enriched populations in comparison to that of whole spleen cells, thymocytes and lymph node cells. The results indicated that NK cell—derived Ly—5 is more similar in migration on SDS-PAGE to T cells and macrophages than B cells, in agreement with current theories on the cell lineage of NK cells (Habu and Okumura, 1982). Studies of the various facilitators and inhibitors of NK cell function have led to a hypothetical model of the NK cell lytic mechanism (Hiserodt, et al, 1983), which ascribes certain properties to the putative NK receptor for target cell interaction. While this model is more fully described in the following sections, one such property would be the susceptibility of the receptor to Cleavage by trypsin-like proteases in a manner that does not completely degrade the fragment generated, allowing it to maintain its interaction with the target cell and participate in subsequent steps of the lytic cycle. In support of this model and the possibility that Ly-5 is a NK cell receptor, the results to be described indicate that trypsin treatment of NK cell—enriched splenocytes generates a fragment of Ly—5 that resists further degradation by trypsin and retains its ability to bind anti-Ly-S mAb. When viewed with data presented on the role of Ly—S in target cell binding and cytotoxicity, the results further implicate Ly-S as a NK cell receptor, in that: 1. Ly-S mediates binding of NK cells to YAC-l tumor cells; 2. Ly-5 has an external protease-sensitive site that is cleaved by gentle trypsinization; 3. the remainder of the external domain fragment generated by gentle trypsinization resists further degradation by high concentrations of trypsin, and; 4. the trypsin-generated fragment maintains its ability to bind to anti-Ly-S mAb, a likely requirement for continued binding to the target cell for the programming and killer cell-indepedent lysis steps of NK cell-mediated cytolysis. While additional activities of the Ly-S fragment need to be demonstrated to confirm its importance as a NK receptor, the results described in this dissertation strongly suggest an important and fundamental role for Ly-S glycoprotein in the natural killer cell system. LITERATURE REVIEW Natural killer (NK) cells are lymphocytes characterized in part by their ability to bind and lyse certain neoplastic cells without prior sensitization. Their discovery came in the process of investigating mechanisms of cell-mediated cytotoxic reactivity against tumors, when Rosenberg, et al, (1972) found that lymphocytes from normal individuals who were free of cancer or known exposure to cancer patients had cytotoxic activity against leukemic cells. Other studies (McCoy, et a1, 1973; Oldham and Herberman, 1973; Takasugi, et a1, 1973; Rosenberg, et al, 1974) showed similar reactivity of lymphocytes from healthy individuals against tumor-derived cell lines. Natural cell-mediated cytotoxicity was subsequently demonstrated to occur in normal rats (Nunn, et al, 1973) and mice (Kiessling et al, 1975a). Splenic natural killer cell activity in mice and rats is an age-dependent phenomenon, appearing at approximately three weeks of age, peaking at five to eight weeks and declining to negligible levels by the age of twelve weeks (Nunn, et a1, 1973; Kiessling, et al, 1975b). However, Imhen peripheral blood lymphocytes (PBL) are examined, murine NK cell development more closely resembles that seen in humans, with significant and stable circulating levels of NK activity maintained for most of the adult life of the mouse (Herberman and Holden, 1978; Lanza and Djeu, 1982). Natural killer cells are nonphagocytic, nonadherent cells that comprise 1—5% of splenocytes and peripheral blood lymphocytes (Kiessling, et al, 1975b; Herberman, et a1, 1975). Their activity is highest in blood and spleen, lower in lymph nodes, lungs, gut-associated lymphoid tissue and bone marrow, and generally absent from thymus (Reynolds, et al, 1981). Using the bone seeking isotope, 89-Sr, or by inducing osteopetrosis with estradiol, NK cell differentiation has been shown to be marrow—dependent (Haller, et a1, 1977; Seaman, et al, 1979; Lust, et a1, 1981, Kumar, et al, 1982). In humans, NK cells have a characteristic morphology on Giemsa staining, with a relatively high cytoplasmic-to-nuclear ratio, kidney-shaped nuucleus, pale cytoplasm and azurophilic granules which has led to their description as large granular lymphocytes (LGL) (Timonen, et al, 1979). Although somewhat smaller, murine NK cells also appear to possess the same LGL morphology and preferentially sediment at 55-60% interfaces on Percoll discontinuous density gradient centrifugation (Tagliabue, et al, 1982). Many of the current theories on the nature, function and cellular origin of natural killer cells have resulted from the characterization of surface antigens on these cytotoxic effectors. Like most murine cells of lymphopoietic or myelopoietic origin, NK cells express the high molecular weight glycoprotein, Ly-5 (Pollack, et a1, 1979; Scheid and Triglia, 1979; Cantor, 1979). Approximately 50% of NK cells bear low levels of Thy-1, which is Characteristic of T lymphocytes (Herberman, et a1, 1978). Thymocytes and NK cells are surface Ly-11+, while mature T cells and cytotoxic T lymphocytes are not (Meruelo, et al, 1980). In common with macrophages is the expression on NK cells of MAC-l and Mph—1.2 antigens (Springer, et a1, 1978; Archer and Davies, 1974), as well as asialo GMl (Kasai, et al, 1980; Okumura, 1982). While essentially all human NK cells highly express receptors for the Fc portion of IgG (FCXR), only a few receptors are identified on murine NK cells and receptors for complement are lacking (Herberman, et al, 1975; Herberman, et al, 1977). Murine NK cells bear Qa-4 and Qa-5 on their cell surfaces and selectively express the NK—l antigen, but lack the B cell antigens IgM and Lyb-2 (Koo, et a1, 1980; Glimcher, et al, 1977). Taken together, the data from comparative surface phenotyping does not allow a designation of cell lineage to NK cells. In order to further Characterize natural killer Cells, describe their function and gain insight into their 063.1 lineage, it is convenient to compare NK—mediated cY1:otoxicity with other forms of cellular cytotoxicity. Natural cytotoxic (NC) cells share with NK cells the ability to lyse certain tumor cells without prior sensitization, but their cytotoxicity is directed toward adherent tumor cell lines derived from solid tumors, while NK cells are preferentially cytotoxic to lymphoid cell lines (Stutman, et al, 1981). The strain distribution of NK and NC activities differ as well, with certain strains of mice having high NC and low NK activity, and vice versa (Stutman, et al, 1978). Stutman, et al (1978), also showed that murine splenic NC activity appears earlier than NK activity, and unlike the latter persists at high levels in aged animals. The kinetics of the cytotoxic response was also different in the two cell types, with significant NK activity appearing within 4 hours, while NC activity required a minimum of 12-18 hours to be observable (Paige, et al, 1978). Differential sensitivity to blocking by simple sugars has also been Observed, with NC activity blocked by the addition of D-mannose, while NK activity is inhibited by a variety of simple sugars (Stutman, et al, 1980). Characterization of cell surface phenotype by treatment with antibody and complement has shown that murine NK cells express asialo GMl (Young, et al, 1980), QaS (Pollack, et al, 1979), Ly—ll (Meruelo, et al, 1980) and NK—l antigens (Burton, 1980), whereas NC cells appear to express only low levels 0f asialo GMl but none Of the other surface antigens (Léittime, et al, 1981). Finally, Djeu, et a1 (1983), have LE 1 10 shown that interleukin-3 (IL-3) allows the selective growth of NC but not NK cells or their precursors, while Henney, et a1 (1981), demonstrated that interleukin-2 (IL-2) augments NK activity and supports the long-term culture of NK cells in yitrg. It is of note that in yitrg NK activity declines rapidly at 37°C in the absence of IL-2 or other stimulatory substances, while NC activity is stable under these conditions (Herberman, et al, 1975; Djeu, et al, 1983). Unlike the many prOperties that serve to distinguish NK cells from NC cells, killer (K) cells mediating antibody-dependent cellular cytotoxicity (ADCC) Show great similarity to NK cells. Ojo and Wigzell (1978), showed that murine NK activity and ADCC directed against certain targets varied in parallel with age, organ distribution and genotype. In addition, both activities were labile i3 yitrg at 37 °C, were augmented by Corynebacterium parvum injection and blocked by cold target inhibition with either target type. Analysis of human K cells (Perussia, et al, 1979), indicated that these cells express the same cell surface phenotype and large granular lymphocyte (LGL) morphology as NK cells. It appears reasonable to conclude on the basis of available data that 13K and K cell activities are separate functions of a Single effector cell. The phenomenon of lectin-dependent cellular CYtotoxicity (LDCC) is also differentiable from classical 11 NK activity and other forms of cellular cytotoxicity. Bonavida and Bradley (1976), showed in mice that lymphocytes must be presensitized in order to generate LDCC activity. Sensitized CTL can be induced by addition of ConA to exert nonspecific spontaneous lytic activity against many cells, including some NK-susceptible targets (Bradley and Bonavida, 1981). When these studies were extended to include spontaneous LDCC in nonstimutated lymphocytes from high and low NK-active mouse strains (Bonavida, 1982), it was observed that in mice with low NK activity, cytotoxicity against NK-susceptible but not NK-resistant targets was facilitated by the addition of ConA, but not by other lectins capable of stimulating CTL-LDCC, such as phytohemagglutinin (PHA) or wheat germ agglutinin (WGA). This activity has been designated NK-LDCC and was abolished by alpha-methyl-D-mannoside, a ConA inhibitor. No significant enhancement of cytotoxicity was observed in lymphocytes from high NK—active strains. The effector cells in NK-LDCC differed from the classical NK effector cell in surface phenotype, being eliminated by treatment with anti-Thy-1.2 and complement (C) but not by anti-asialo GMI + C. While these data do not allow a precise understanding of the relationship of NK activity to LDCC, the inverse relationship observed among mouse strains between NK activity and NK-LDCC may eventually provide insight into the processes leading to the development of functional natural killer cells. 12 A comparison 0f NK cells and cytotoxic T lymphocytes (CTL) reveals many differences and some similarities. Unlike CTL, NK cells appear to lack any major histocompatability restriction for cytotoxic activity (Herberman and Ortaldo, 1981; Stutman, et al, 1978), evidence of immunologic memory (Dennert, 1980), or the need for priming to induce a cytotoxic response (Herberman and Holden, 1978). Characterization of cell surface phenotype by abrogation of activity following reaction with specific antibody and complement suggests that murine NK cells but not CTL express Ly-11.2, Qa5 and asialo GMl, while CTL but not NK cells express Lyt-2 (Meruelo, et al, 1980, Minato, et al, 1981; Kasai, et al, 1980). While both effector cell types express the hematOpoietic differentiation antigen Ly-5/T200, only NK cell activity is blocked by anti-T200 or anti-Ly-S in the absence of complement (Seaman, et al, 1981; Kasai, et al, 1979). Conversely, anti—RAT* sera appears to block both the binding and post-binding lethal hit stage of CTL-mediated lysis without affecting ADCC or NK activity (Hiserodt and Bonavida,1981). Roder, et a1 (1979), isolated cell surface glycoproteins from NK-susceptible targets that specifically blocked binding of NK cells but not CTL to their respective targets, suggesting that different receptors are involved in the two cytolytic processes. The effect of interferon treatment on target cells differs as well, resulting in increased sensitivity of 13 CTL-susceptible targets but decreased sensitivity of NK-susceptible targets to the respective forms of cytotoxicity (Welsh, et a1, 1981). Under certain circumstances, it appears possible to show both NK and CTL activities in a single cloned lymphocyte population. Brooks (1983), generated two allospecific CTL clones in medium containing 2% ConA-induced spleen cell supernatant, which has been termed conditioned medium (CM). One Clone, AIG7, demonstated H-2b specificity and lysed only the H-2b-bearing target EL4 under the conditions described above. However, after one week culture in 40% CM, AIG7 cells lysed not only EL4 cells but also the NK-susceptible targets YAC-l and 27v-IC2 (L5178Y-lymphoma), while having no activity against the NK-resistant targets P815 (H-2d) and av (L5178Y—lymphoma). This effect was reversible by returning the cultures to 2% CM. Partially purified interferon (IFN) and interleukin-2 (IL-2) preparations individually supported the augmented response, and their effects were synergistic when used in combination, while addition of ConA to 2% CM did not promote NK activity. Similar effects were seen with the other CTL clone studied, suggesting that at least in some circumstances a single cell can give rise to both NK and CTL activities, that CTL are not necessarily terminally differentiated cells and that IFN and IL—2 are mediators of this phenomenon. 14 Determining whether NK and CTL activities result from different cell types or alternate functions of a single effector cell was further complicated by the finding that mitogen- or allogeneic cell-stimulated lymphocytes express not only CTL but also NK and ADCC activities (MacDonald and Bonnard, 1975; Ortaldo, et al, 1977). In addition, the cells responsible for the NK—like activity in these cultures have been shown to bear high surface levels of Thy-l (Dennert, et al, 1981; Brooks, et al, 1982), while only about 50% of unstimulated splenic NK cells express even low levels Of Thy-1 (Herberman, et al, 1978). The cytotoxic activities observed in such cultures led to additional cell designations, such as activated killer (AK), mixed lymphocyte culture-activated killer (MLC-AK), anomolous killer (AK), NK-like effector, NK-T and lymphokine-activated killer cells (reviewed by Timonen, 1983). Seeley, et a1 (1979), have identified several similarities between NK and MLC-AK cells, including low affinity receptors for sheep red blood cells, similar ranges of target cell sensitivity, cross competition between sensitive targets and the lack of an amnestic response or immunologic memory. However, numerous studies exist to suggest significant differences between the various activities. In human mixed lymphocyte cultures (MLC), Zarling and Bach (1982) have shown that most if not all of the cells responsible for CTL activity are OKT3+ and OKT8+ while NK-like cells are OKT3- and OKT8-. Abo 15 and Balch (1982) found evidence that the cell surface phenotoype and kinetics of response differ between NK and MLC-AK cells, with human NK cells expressing the HNK-l antigen and having spontaneous activity while MLC-AK cells lacked this surface marker and required 7-day stimulation with alloantigen or mitogen to acquire cytotoxic activity. MLC-AK cells also appear to lack cell surface receptors for the Fc portion of immunoglobulin (FcR), that are Characteristic of human NK cells (Seeley, et al, 1979). In addition, certain target recognition structures appear to be unique to NK cells (Takasugi, et al, 1977). While several features serve to contrast NK and MLC-AK cells, the latter differ in several respects from allospecific CTL as well. Cytotoxic T lymphocytes display classic memory responses whereas activated killer cells do not; their respective targets also fail to cross compete in cold target inhibition studies (Seeley, et a1, 1979). Seeley and Golub (1978), studied the kinetics of deveIOpment of the two responses during in yitrg culture and found that peak activity of MLC-AK cells occured at 4-5 days and returned to baseline by day 9, while CTL activity peaked at 6—7 days and remained active for months. Further understanding of the role of the various cytotoxic effector cells to one another is expected as additional clones of cells posssessing one or more of these cytotoxic activities become available for study. 16 Studies into the mechanism of natural killer cell-mediated cytotoxicity and its regulation have shown that cytotoxicity results from a multiple—step process. Roder and Kiessling (1978), identified a subpopulation of nylon wool nonadherent murine splenocytes that bound specifically to NK—sensitive target cells, suggesting that target cell binding was an important step in the cytolytic process. They went on to show that: binding preceeded lysis and was inhibited by EDTA; the tumor—binding cells (TBC) resembled small lymphocytes with membrane specialization in areas of tumor cell contact; metabolic inhibitors and serine protease inhibitors blocked lysis without affecting target cell binding and interferon—inducing agents augmented lysis without increasing the number of TBC (Roder, et al, 1978). Data from other laboratories has confirmed that cytotoxicity requires binding between NK effectors and target cells, is temperature-independent and Mg++-dependent (Carpen, et al, 1981; Quan, et al, 1982; Hiserodt, et al, 1982). The Ca++ pulse method, in which effector-target cell conjugates are held in medium with Mg++/EGTA and CaC12 subsequently added, served to establish that binding alone is not sufficient for activation of the lytic mechanism in NK cell—mediated cytolysis and that activation and programming for lysis is Ca++-dependent and requires energy (Hiserodt, et al, 1982; Quan, et al, 1982). Saksela, et al (1982), employed fluorochrome—coupled 17 lectins that bind preferentially to the golgi apparatus of fixed and detergent-solubilized cells to show that programming for lysis is associated morphologically with movement of the golgi apparatus toward the target cell. They also observed that the resultant lytic phase of NK-target interaction was inhibited by monensin, a carboxylic ionophore that specifically blocks vesicular traffic of golgi derived vacuoles to the cell surface (Tartakoff, 1981), in support of the stimulus secretion model of NK cytolysis proposed by Roder, et a1 (1980). After Ca++—dependent programming, the addition of EDTA to NK-target conjugates followed by dispersion into dextran containing medium was shown to dissociate all conjugates, but not interfere with target cell lysis as determined by 51-Cr release, thereby confirming the existence of a killer cell-independent lysis (KCIL) step in the overall cytotoxic process (Hiserodt, et al, 1982). KCIL was found to be Ca++ and Mg++-independent but protease and antiprotease sensitive (Hiserodt, et a1, 1983; Hudig, et al, 1981). After detaching from the KCIL—programmed target, the NK cell is apparently refractory to further cell binding or lysis for a period of time, in which it is said to be recycling (Timonen, T., et al, 1982; Ullberg and Jondal, 1981). The existence of KCIL and evidence for a stimulus secretion model of NK cytolysis supported a search for soluble mediators of the cytotoxic process. Using effector cells incubated with l8 PHA or ConA in the bottom of a miniature Marbrook chamber, Wright and Bonavida (1981), demonstrated that the effector cells released soluble mediators which traversed a 0.2 micron Nucleopore membrane barrier to specifically lyse NK-sensitive target cells cultured in the top of the chamber. They also showed that the mediators, termed natural killer cell—derived cytotoxic factors (NKCF), were released from spleen cells but not thymocytes, had a molecular weight greater than 12,000 daltons, were stable at -20 °C, inactivated by heating to 100°C for 15 minutes, and inhibited in their activity by the addition of alpha—methyl-D-mannoside, D-galactose, or N-acetyl-D—galactosamine, which distinguishes them from lymphotoxin activity against the same targets (Toth, and Granger, 1979). However, unlike the rapid kinetics of cellular NK activity in vitro, NKCF-induced cytotoxicity required incubation periods of 40 hours to cause significant target cell lysis as measured by trypan blue exclusion (Wright and Bonavida, 1981; Farram and Targan, 1983), suggesting that additional effector cell-associated components may be necessary for full expression of cytotoxic activity. The finding that proteases inhibited both target cell binding and lysis while antiproteases blocked only lysis (Roder, et al, 1978), implicated a protein with protease activity in the lytic step of NK-mediated cytolysis. Hudig, et a1 (1981), demonstrated that 19 alpha-l-antichymotrypsin is a potent inhibitor of NK activity and similar results were shown for alpha-Z-macroglobulin, a potent inhibitor of many endOpeptidases including chymotrypsin (Gravagna, et al, 1982; Laurell and Jeppson, 1975). In detailed studies of protease inhibition of the KCIL stage of human NK cytolysis, Hiserodt, et al (1983a, 1983b), found that trypsin (T), chymotrypsin (CT), and papain (P) inhibited lysis well beyond the Ca++-dependent programming step. However, in an NKCF assay only T and CT but not P inhibited cytolysis. Preincubation of target cells with T, CT or P failed to diminish their ability to absorb NKCF or to serve as cold target competitors in an NKCF assay, but abolished their ability to serve as targets in a standard NK assay due to decreased binding by effector cells. Pretreatment of effector cells with T and CT but not P blocked their abiltiy to mediate NK cytolysis. The authors concluded that the receptor on target cells for binding to the effector was T-, CT- and P-sensitive, the receptor for NKCF was protease-resistant and that both NKCF and the NK receptor for binding to target antigen were T- and CT-sensitive but P-resistant (Hiserodt, et al, 1983b). Taken together with data from other studies, the authors proposed a multistage hypothetical model for human natural killing, in which: 1. Mg++-dependent binding of NK cell surface receptor to target cell antigen occurs, followed by; 2. Ca++-dependent binding to an 20 extracellular site analogous to that seen in CTL cytolytic mechanisms (Martz, et al, 1982): 3. enzymatic cleavage of a component of the NK surface receptor (lytic focusing fragment), which remains attached to the target cell and has otherwise cryptic high avidity subunits for NKCF that are exposed in the process of cleavage; 4. intracellular flux of Ca++ through calcium Channels to activate the more protracted phase of Ca++-dependent programming. These channels can be blocked by the calcium channel blocking agent verapamil (Hiserodt, et al, 1982a); 5. secretion of NKCF; 6. focusing and polymerization of NKCF by the target cell-bound lytic focusing fragment, possibly through disulfide bond formation, since 2—mercaptoethanol additon after the Ca++—dependent step inhibits cytotoxicity (Hiserodt, et al, 1982b), and; 7. interaction of the assembled lytic complex with the target cell membrane in a manner analogous to the membrane attack complex of the complement system, leading to colloid osmotic lysis. This hypothetical model serves as a framework for the inquiry into potential mediators of the various stages of the cytolytic process as well as testing of the hypothesis itself. Evidence has accumulated in recent years to suggest a role for natural killer cells in host defense mechanisms against cancer (Herberman and Holden, 1978a; Hanna and Fidler, 1980; Talmadge, et al, 1980; Warner and Dennert, 1982), viral, fungal and parasitic infections (Welsh, 20 extracellular site analogous to that seen in CTL cytolytic mechanisms (Martz, et al, 1982): 3. enzymatic cleavage of a component of the NK surface receptor (lytic focusing fragment), which remains attached to the target cell and has otherwise cryptic high avidity subunits for NKCF that are exposed in the process of cleavage; 4. intracellular flux of Ca++ through calcium Channels to activate the more protracted phase of Ca++-dependent programming. These channels can be blocked by the calcium channel blocking agent verapamil (Hiserodt, et al, 1982a); 5. secretion of NKCF; 6. focusing and polymerization of NKCF by the target cell-bound lytic focusing fragment, possibly through disulfide bond formation, since 2-mercaptoethanol additon after the Ca++-dependent step inhibits cytotoxicity (Hiserodt, et al, 1982b), and; 7. interaction of the assembled lytic complex with the target cell membrane in a manner analogous to the membrane attack complex of the complement system, leading to colloid osmotic lysis. This hypothetical model serves as a framework for the inquiry into potential mediators of the various stages of the cytolytic process as well as testing of the hypothesis itself. Evidence has accumulated in recent years to suggest a role for natural killer cells in host defense mechanisms against cancer (Herberman and Holden, 1978a; Hanna and Fidler, 1980; Talmadge, et al, 1980; Warner and Dennert, 1982), viral, fungal and parasitic infections (Welsh, 21 1979; Murphy and McDaniel, 1982; Hatcher and Kuhn, 1982), in graft rejection (Kiessling, et al, 1977; Datta, et al, 1979; Warner and Dennert, 1982), and even in the regulation of normal hematopoietic cell differentiation (Abruzzo and Rowley, 1983). This broad spectrum of activity has also raised interest in the possible immunotherapeutic use of NK cells in the treatment of disease. Of particular interest in this regard is the identification of physiologic and pharmacologic regulators of NK cytolytic activity, as well as the development of functional NK cells from precursor cells. Interferon has important but mixed effects on natural killer cell activity against susceptible tumor targets. Interferon rapidly boosts NK activity (Roder, et al, 1978; Herberman, et al, 1980; Ortaldo, et al, 1982), apparently by increasing the number of TBC that lyse their targets (one way of defining an NK precursor cell), accelerating lysis and recycling capacity but not by increasing the number of effectors that bind target cells (Roder, et al, 1978; Ullberg and Jondal, 1981; Targan and Stebbing, 1982). One exception to this is the ability of interferon to increase the number of TBC against anchorage-dependent tumor and nontumor targets in humans (Timonen, et al, 1982). In addition to their ability to respond to interferon with increased cytolytic activity, NK cells were shown to be capable of producing interferon in response to target cell contact in a positive feedback mechanism 22 (Timonen, et al, 1980). While interferon is known to have direct inhibitory effects on tumor cells (Gresser, et al, 1970), it also seems capable of protecting certain normal and malignant cells from NK-mediated cytolysis when such cells are incubated with interferon prior to contact with NK cells (Welsh, et al, 1981), although such treatment increases their susceptibility to CTL-mediated cytotoxicity. Trinchieri, et al (1981), showed that normal cells were more rapidly and effectively protected from NK cytolysis by interferon than tumor cells, suggesting a mechanism for the selective destruction of tumor cells in yiyg. Interleukin-2 (IL-2, T cell growth factor, TCGF), is an potent augmentor of NK activity and is the only growth factor detected to date that is capable of maintaining the long-term proliferation of cytolytically active NK cells (Henney, et al, 1981). In fact, NK cultures and clones have been generated and evaluated using IL-2 and interferon-containing conditioned medium (CM) supernatants obtained from splenocyte cultures stimulated With the mitogenic lectins PHA or ConA (Nabel, et al, 1981a and 1981b; Brooks, et al, 1982). IL—2 binding studies of NK cells cultured in the presence or absence of interferon has led Kuribayashi, et al (1981), to suggest that a synergistic effect of the two lymphokines may occur, with interferon causing an increase in the number or affinity of IL-2 receptors on NK cells as part of the augmentation 23 process. Studies investigating the cell surface phenotype of cells responding to interferon and IL-2 with increased cytotoxicity has led to the description of subsets of NK cells (Minato, et al, 1981; Minato and Bloom, 1982; Burton, et al, 1982). Based on residual cytotoxic activity of murine spleen and marrow cells following treatment with specific antibody and complement, as well as the response to IL-2 and interferon preparations, Minato, et al (1981), identified four phenotypically different cytotoxic cell populations: 1. NK-I cells, prominent in spleen, Thy-1-, Qa5+, Ly-5+, augmented by interferon; 2. NK-T cells, prominent in spleen, Thy-1+, Qa5+, Ly-5+, augmented by interferon and IL—2; 3. NK-M cells, restricted to marrow, Thy-1-, Qa5-, Lyt—2-, Ly-5+, not modulated by IFN or IL-2, and; 4. TK cells, in spleen, Thy-1+, Qa5-, Lyt-2+, Ly-5+, augmented by IL-2 but not IFN, lyse CTL—sensitive target but not NK-sensitive target. These results and others employing complement-mediated lysis to determine cell surface phenotype must be interpreted with caution, since surface marker expression does not always indicate susceptibility to antibody + C (Woody, et al, 1977; Dennert, et al, 1980). Even so, the probable existence of heterogeneity within the natural killer cell system and the identification of relevant surface antigens has served to guide the study of NK cell development and lineage. 24 Despite the abundance of literature regarding the nature of natural killer cells, their function and relationship to other immunologic effectors, relatively little is currently known about precursor or stem cells that give rise to NK cells. One approach to this problem involves the examination of cell surface antigens expressed on differentiated cells but lacking on their immature counterparts. In addition, if such a surface antigen is also implicated in an important aspect of cellular function, it represents a unique Opportunity to explore the developmental biology and biochemistry of functional cellular maturation. Such may be the case with the murine lymphopoietic and myelopoietic cell surface differentiation glycoprotein, Ly-5. Several studies have shown that polyvalent anti-Ly-S sera block NK activity, even in the absence of complement (Kasai, et al, 1979; Pollack, et al, 1979; Minato, et al, 1980), suggesting that the Ly-5 surface molecule may directly participate in one or more steps of NK-mediated cytolysis. After reports Showed that Ly-5 and T200 antigens are expressed on identical murine cell surface proteins (Siadak and Nowinski, 1980), and that Ly—5 is polymorphic (Komuro, et al, 1975; Omary, et al, 1980), while T200 is not (Trowbridge, et al, 1975; Trowbridge and Mazauskas, 1976), Seaman, et al (1981), found that anti—T200 monoclonal antibody (mAb) blocked NK cytolysis without affecting CTL activity against the same target. In 25 addition, they observed that pretreatment of YAC-l target cells with anti-T200 did not alter their susceptibility to NK cytotoxicity even though by immunofluorescence these targets express T200, and that anti-T200 mAb must be continuously present in the NK cell assay for inhibitory activity to be observed. This requirement may help to explain the variable effects of anti-Ly-S sera on NK activity observed in some experiments (Minato, et al, 1980). In addition, some evidence exists to suggest that different subsets of NK cells may vary in their susceptibility to inhibition by anti-Ly-S sera, with NK—I cells blocked more effectively than NK-M cells (Minato, et al, 1981). Direct antibody binding studies to determine the distribution of Ly-5/T200 on hematOpoietic cells have revealed that Ly-5+ cells include all known sets of T cells, prothymocytes, myeloid cells, monocytes, macrophages, natural killer cells and subsets of B cells, while Ly-5- cells include erythrocytes, proerythroblasts, hepatocytes and kidney cells (Komuro, et al, 1975; Scheid and Triglia, 1979). Susceptibility to lysis by anti-Ly-S + C has also been used to estimate the percentage of lymphocytes that express surface Ly—S and determine their biological function (Minato, et al, 1980, but several studies have indicated that surface Ly-5 positivity does not necessarily correlate with the ability to be lysed with antibody and complement. Minato, et al (1980), found 26 that anti-Ly-S sera + C lysed 30-40% of CBA/J and 7% of BALB/c nude mouse splenocytes and abrogated NK activity, leading them to conclude that such were the percentages of Ly-5+ lymphocytes in the respective mouse strains and that all mature Ly-5+ cells were lysed by this technique. Following a three hour treatment with an interferon-containing preparation, they identified recovery of NK activity and additional cells susceptible to lysis by anti-Ly-S sera + C, from which they concluded that Ly-S- precursors of NK cells exist in murine spleen, and that these cells are induced by interferon to convert to Ly-5+ cells with natural killer cell function. However, using a second step antibody (anti-rat Ig), Seaman, et al (1981), showed that anti-T200 mAb lysed 82% of C57BL/6 splenocytes, suggesting that additional Ly-5+ cells may have been present in CBA/J and BALB/C nude spleens that resisted complement-mediated lysis. Further supportive evidence came from Dennert, et al (1980), who found that in the absence of a second step antibody, anti-T200 mAb lysed more than 95% of C57BL/6 thymocytes, but only 40% of spleen cells, 20-30% of bone marrow cells and 5-10% of BALB/C nude spleen cells, in close aggreement with the results of Minato, et al (1980). However, the former investigators also looked at the effect of anti-T200 + C on CFU—S and prothymocytes and found that neither were diminished by this treatment. In addition, they subjected nonerythrocytic thymic, splenic and bone 27 marrow cells to quantatative analytical cytofluorimetry and found that 78% of splenocytes, 91% of thymocytes and 81% of bone marrow cells were Ly-5+, verifying that a large percentage of Ly-5+ cells exist in bone marrow and spleen that resist lysis with anti-Ly-S and complement (Dennert, et al, 1980). From these results and the evaluation of CTL activity, helper-T cell function, mixed lymphocyte reactivity, T and B cell mitogenesis and T cell-independent B cell responsiveness, the authors also demonstrated that anti—T200 mAb is selectively cytotoxic for thymocytes and mature thymus-dependent T lymphocytes. Based on the above information, it appears unsettled as to whether splenic Ly-5- precursor cells actually convert to Ly—5+ NK effector cells as proposed by Minato, et al (1980), since it may be equally possible that Ly-5+ NK cells could have been blocked in their cytolytic capacity by anti—Ly-S sera but were resistant to complement-mediated lysis, following which the cells regenerated additional cell surface Ly-S or other relevant molecules in the presence of interferon or other lymphokine activity to regain cytolytic function. In addition, the incremental increase in cells susceptible to lysis by anti-Ly-S + C after interferon treatment may have represented a cellular maturation event akin to that of prothymocytes to thymocytes and been unrelated to new NK effectors generated from precursors. In order better understand NK cell development and its regulation, 28 additional information is needed regarding the role of Ly-S in the mechanism of natural killer cell-mediated cytotoxicity and alternate approaches developed to determine the cell surface phenotype of precursors and effectors. Before proceeding to describe the biochemistry and heterogeneity of -the Ly-5 system, it is useful to establish the relevance of this system to human natural killer cells. Omary, et al (1980), identified the human homologue of murine T200 glycoprotein and documented the following similarities between the two systems: 1. the peptide structure of the glyCOprotein is highly conserved between the two species, a property shared by only a few other cell surface proteins such as Ia, Ig, Thy—1 and MHC antigens; 2. the molecule is broadly distributed within the hematopoietic system but not detectable on nonhematopoietic cells, and; 3. structural differences between the forms of the glycoprotein are found on T and B lymphoblastoid cell lines. Sparrow and McKenzie (1983), showed that anti-human T200 mAb specifically blocks binding and NK-mediated lysis of K562 target cells without affecting CTL responses. Targan and Newman (1983), presented evidence to suggest that T200 may also participate in the early part of the Ca++—dependent programming step of NK lysis, in that 13.1 mAb, which recognizes an epitope on T200, blocks lysis if added prior to Ca++ and does so without altering conjugate formation. 29 Together, these data strongly suggest an important and fundamental role for Ly-5/T200 in the natural killer cell system and establish the murine system as a reasonable model for exploring the nature and function of the Ly-5 glyCOprotein. Any attempt to ascribe biological significance to the Ly-5 glycoprotein must contend with the fact that the expression of Ly-5 is not unique to NK cells. Early evidence that differences in Ly-5 may exist among various lymphoid cell populations came from Trowbridge, et al (1975), who found that the glyCOprotein migrates on sodium docecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), with a relative molecular mass (Mr) of 170,000—190,000 for thymocytes and T cells, while B cells express a 220,000 form. Similar Mr were estimated by Dunlap, et al (1978). Michaelson, et a1 (1979), found three peaks for splenocytes, at 175, 185 and 220 kilodaltons (KD), while thymocytes had a 175 KD peak which was broad, typical of heavily glycosylated proteins. Coffman and Weissman (1981), developed a monoclonal antibody that reacted specifically with B cells, precipitating the 220,000 dalton form of T200, which they designated B220. Tung, et al (1981), studied the Ly-5 expression of 7 cloned cell lines and found that each expressed only one of three forms: T cells, 200,000; macrophages, 205,000 and; B cells, 220,000. Cleveland peptide maps for the three molecular weight forms were 30 identical, suggesting that differences in the molecular weights of the various forms of Ly-5 may reside in the carbohydrate portion of the glycoprotein. SDS-PAGE analysis of Lyt-2,3+ CTL Clones revealed additional apparent forms of Ly-5 with Mr of 210,000 and 215,000 (Tung, et al, 1983). The relative molecular mass of Ly-5 that is expressed on NK cells may provide insight into their cell lineage but has yet to be Characterized. Trowbridge, et al (1977), identified T200 as a cell surface protein that is partially degraded by gentle trypsinization. Saturating binding experiments suggest that murine splenocytes express approximately 20,000 T200 molecules per cell, while thymocytes express slightly more, marrow cells slightly less and certain lymphomas up to 5 times as much of this glycoprotein (Trowbridge, 1978). Omary and Trowbridge (1980), further characterized T200 as a phosphorylated transmembrane glyCOprotein with a relatively protease-resistant domain exposed on the cell surface and containing the antigenic determinant recognized by anti-T200 mAb. Trypsinization of radioiodinated, Nonidet-P40 (NP-40) solubilized cell surface proteins from a murine T cell lymphoma line (BW5147), showed that the protease-resistant fragment migrated with a Mr of 100,000 on SDS—PAGE, contained all of the sites labeled by lactoperoxidase (LPO)-catalyzed radioiodination of viable cells and most if not all of the mannose-containing oligosaccharide units of the T200 31 glycoprotein. The remainder of the protein was extensively degraded by trypsin and contained phosphoserine residues that could be labeled with 32-PO4 in yiyg. The existence of a protease-resistant outer domain in a molecule which has been shown to be important in NK-mediated cytolysis, a process that involves proteases, makes Ly-S/TZOO a candidate molecule for facilitating the binding and/or lytic stages of natural killer cell cytotoxic reactivity against susceptible tumor cells. Watson, et al (1981), looked at intracytoplasmic forms of Ly-5/T200 in murine T and B cells and found that EL-4 cells (a T cell leukemia) produced a 170 KD precursor that reacted with anti—Ly-S sera and was not expressed on the cell surface, while the corresponding precursor form in LPS blasts (B cells) had a molecular weight of 190 KD. Precursors were shown to have higher 3H-mannose:3H-glucosamine ratios than their immediate products and lacked the 32-PO4 incorporation characteristic of the mature forms of the glyCOprotein. Investigation of the oligosaccharide portion of human Ly-5/T200 glycoprotein (also known as human leukocyte common antigen), has revealed that it labels with 3H-mannose, 3H—galactose and 3H-fucose (van Eijk and Muhlradt, 1981), carries 0- and N-glycosidically linked, sialated carbohydrate chains and that the O—linked Chains of B cells but not T cells are highly susceptible to degradation by endo-beta-galactosidase (Childs, et al, 32 1983). Using neuraminidase treatment of human B and T cell lines, Morishima, et al (1982), concluded on the basis of mobility changes on SDS-PAGE that the lower molecular weight form of Ly—5/T200 seen in T cell lines had higher sialic acid content than the higher molecular weight form in B cell lines and are therefore more highly glycosylated or have greater chain branching. The oligosaccharide structure of murine Ly—5 glycoprotein is as yet undetermined. LITERATURE CITED Abo, T., and C. M. Balch. 1982. Characterization of HNK-1+ (Leu-7) human lymphocytes. II. Distinguishing phenotypic and functional properties of natural killer cells from activated NK-like cells. J. Immunol. 129:1758. Abruzzo, L. V., and D. A. Rowley. 1983. Homeostasis of the antibody response: immunoregulation by NK cells. Science 222:581. Archer, J. R., and A. L. Davies. 1974.. Demonstration of an alloantigen on the peritoneal exudate cells of inbred strains of mice and its association with Chromosome 7 (linkage group I). J. Immunogen. 1:113. Bonavida, B., and T. P. Bradley. 1976. Studies on the induction and expression of T-cell mediated immunity. V. Lectin-induced nonspecific cell-mediated cytotoxicity by alloimmune lymphocytes. Transplantation. 21:94. Bonavida, B. 1982. Lectin-dependent natural killer cellular cytotoxicity in mice (NK-LDCC): a new suprpulation of NK-like cytotoxic cells. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 483. Bradley, T. P., and B. Bonavida. 1981. Mechanism of cell- mediated cytotoxicity at the single cell level. III. Evidence that cytotoxic T lymphocytes lyse both antigen-specific and -nonspecific targets pretreated with lectins or periodate. J. Immunol. 126:208. Brooks, C. G., K. Kuribayashi, M.-F. Feng, and C. S. Henney. 1982. Characterization of cloned murine cell lines having high cytolytic activity against YAC-l targets. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 851. Brooks, C. G. 1983. Reversible induction of natural 33 34 killer cell activity in cloned murine cytotoxic T lymphocytes. Nature 305:155. Burton, R. C. 1980. Alloantisera selectively reactive with NK cells: Characterization and use in defining NK cell classes. In Natural Cell-Mediated Immunity against Tumors. Edited by R. B. Herberman. Academic Press, New York. 19. Burton, R. C., S. P. Bartlett, and H. J. Winn. 1982. Alloantigens specific for natural killer cells. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 105. Cantor, H., M. Kasai, F. W. Shen, J. C. Leclerc, and L. Glimcher. 1979. Immunogenetic analysis of ‘natural killer' activity in the mouse. Immunol. Rev. 44:3. Carpen, O., I. Virtanen, and E. Saksela, 1981. The cytotoxic activity of human natural killer cells requires an intact secretory apparatus. Cellular Immunol. 58:97. Childs, R. A., R. Dalchau, P. Scudder, E. F. Hounsell, J. W. Fabre, and T. Feizi. 1983. Evidence for the occurence of O-glycosidically linked oligosaccharides of poly-N-acetyllactosamine type on the human leukocyte antigen. Biochem. Biophys. Res. Comm. 110:424. Coffman, R. L. and I. L. Weissman. 1981. B220: a B cell- specific member of the T200 glycoprotein family. Nature, 289:681. Datta, S. K., M. T. Gallagher, J. J. Trentin, R. Kiessling, and Wigzell, H. 1979. Apparent identity of mechanisms of genetic resistance to marrow transplantation and natural killer cell activity. Biomedicine 31:62. Dennert, G. 1980. Cloned lines of natural killer cells. Nature 287:47. Dennert, G., R. Hyman, J. Lesley, and I. S. Trowbridge. 1980. Effects of cytotoxic monoclonal antibody specific for T200 glycoprotein on functional lymphoid cell populations. Cell. Immunol. 53:350. Dennert, G., G. Yogeeswaran, and S. Yamagata. 1981. Cloned cell lines with natural killer activity: Specificity, function, and cell surface 35 markers. J. Exp. Med. 153:545. Djeu, J., E. Lanza, S. Pastore, and A. J. Hapel. 1983. Selective growth of natural cytotoxic but not natural killer effector cells in interleukin-3. Nature. 306:788. Dunlap, B., F. H. Bach, and M. L. Bach. 1978. Cell surface changes in alloantigen activated T lymphocytes. Nature. 271:253. Farram, E. and S. R. Targan. 1983. Identification of human natural killer soluble cytotoxic factor(s) (NKCF) derived from NK-enriched lymphocyte populations: specificity of generation and killing. J. Immunol. 130:1252. Glimcher, L., F. W. Shen, and H. Cantor, H. 1977. Identification of a cell-surface antigen selectively expressed on the natural killer cell. J. Exp. Med. 145:1. Gravagna, P., E. Gianazza, P. Arnaud, M. Neels, and E. W. Ades. 1982. Modulation of the immune response by plasma protease inhibitors. 11. Alpha-Z—macroglobulin subunits inhibit natural killer cell cytotoxicity and antibody—dependent cell-mediated cytotoxicity. Scand. J. Immunol. 15:115. Gresser, I. D., D. Brouty-Boye, M. T. Thomas, and A. Marcieira-Coelho. 1970. Interferon and cell division. I. Inhibition of the multiplication of mouse leukemia L1210 cells in vitro by interferon preparations. Proc. Natl. Acad. Sci. 66:1052. Haller, O., R. Kiessling, A. Orn, and H. Wigzell. 1977. Generation of natural killer cells: an autonomous function of the bone marrow. J. Exp. Med. 145:1411. Hanna, N. and I. J. Fidler. 1980. Role of natural killer cells in the destruction of Circulating tumor emboli. J. Natl. Cancer Inst. 65:801. Hatcher, F. M. and R. E. Kuhn. 1982. Natural killer (NK) cell activity against extracellular forms of Trypanosoma cruzi. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 1091. Henney, C. S., K. Kuribayashi, D. E. Kern, and S. Gillis. 36 1981. Interleukin-2 augments natural killer cell activity. Nature. 291:335. Herberman, R. B., M. E. Nunn, H. T. Holden, and D. H. Lavrin. 1975. Natural cytotoxic reactivity of mouse lymphoid cells against syngeneic and allogeneic tumors. II Characterization of effector cells. Int. J. Cancer. 16:230. Herberman R. B., S. Bartram, J. S. Haskill, M. Nunn, H. T. Holden, and W. H. West. 1977. Fc receptors on mouse effector cells mediating natural cytotoxicity against tumor cells. J. Immunol. 119:322. Herberman, R. B., and H. T. Holden. 1978. Natural cell-mediated immunity. Adv. Cancer Res. 27:305. Herberman, R. B., J. R. Ortaldo, J. Y. Djeu, H. T. Holden, J. Jett, N. P. Lang, and S. Pestka. 1980. Role of interferon in regulation of cytotoxicity by natural killer cells and macrophages. Ann. N. Y. Acad. Sci. 350:63. Herberman, R. B., and J. R. Ortaldo. 1981. Natural killer cells: their role in defenses against disease. Science. 214:24. Hiserodt, J. C. and B. Bonavida. 1981. Studies on the induction and expression of T cell mediated immunity. XI. Inhibition of the ‘lethal hit' in T cell-mediated cytotoxicity by heterologous rat antiserum made against alloimmune cytotoxic T lymphocytes. J. Immunol. 126:256. Hiserodt, J. C., L. J. Britvan, and S. R. Targan. 1982a. Characterization of the cytolytic reaction mechanism of the human natural killer (NK) lymphocyte: resolution into binding, programming and killer cell-independent steps. J. Immunol. 129:1782. Hiserodt, J. C., L. J. Britvan, and S. R. Targan. 1982b. Differential effects of various pharmacologic agents on the cytolytic reaction mechanism of the human natural killer lymphocytes: further resolution of programming for lysis and KCIL into discrete steps. J. Immunol. 129:2266. Hiserodt, J. C., L. J. Britvan, and S. R. Targan. 1983a. Studies on the mechanism of the human natural killer cell lethal hit: evidence for transfer of protease-sensitive structures requisite for target cell lysis. J. Immunol. 131:2710. 37 Hiserodt, J. C., L. J. Britvan, and S. R. Targan. 1983b. Studies on the mechanism of the human natural killer cell lethal hit: analysis of the mechanism of protease inhibition of the lethal hit. J. Immunol. 131:2705. Hudig, D., T. Haverty, C. Fulcher,, D. Redelman, and J. Mendelsohn. 1981. Inhibition of Human natural cytotoxicity by macromolecular antiproteases. J. Immunol. 126:1569. Kasai, M., J. C. Leclere, F.-W. Shen, and H. Cantor. 1979. Identification of Ly 5 on the surface of ‘natural killer' cells in normal and athymic inbred mouse strains. Immunogenetics. 8:153. Kasai, M., M. Iwamori, Y. Nagai, K. Okumura, and T. Tada. 1980. A glycolipid on the surface of mouse natural killer cells. Eur. J. Immunol. 10:175. Kiessling, R., E. Klein, and H. Wigzell. 1975a. ‘Natural' killer cells in the mouse. 1. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Specificity and distribution according to genotype. Eur. J. Immunol. 5:112. Kiessling, R., E. Klein, H. Pross, and H. Wigzell. 1975b. ‘Natural' killer cells in the mouse. II. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Characterization of the killer cell. Eur. J. Immunol. 5:117. Kiessling, R., P. S. Hochman, O. Haller, G. M. Shearer, H. Wigzell, and G. Cudkowicz. 1977. Evidence for a similar or common mechanism for natural killer cell activity and resistance to hemopoietic grafts. Eur. J. Immunol. 7:655. Komuro, K., K. Itakura, E. A. Boyse, and M. John. 1975. Ly-5: a new T—lymphocyte antigen system. Immunogenetics. 1:452. Koo, G. C., J. B. Jacobson, G. J. Hammerling, and U. Hammerling. 1980. Antigenic profile of murine natural killer cells. J. Immunol. 125:1003. Kumar, V., P. F. Mellen, and M. Bennet. 1982. Marrow dependence of natural killer cells. In NK Cells and Other Natural Effector Cells. Editied by R. B. Herberman. Academic Press, New York. 329. Kuribayashi, K., S. Gillis, D. E. Kern, and C. S. Henney. 1981. Murine NK cell cultures: effects of 38 interleukin-2 and interferon on cell growth and cytotoxic reactivity. J. Immunol. 126:2321. Lanza, E. and J. Djeu, 1982. Age-independent natural killer cell activity in murine peripheralblood. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 335. Lattime, E. C., G. A. Pecoraro, and O. Stutman. 1981. Natural cytotoxic cells against solid tumors in mice. III: A comparison of effector cell antigenic phenotype and target cell recognition structures with those of NK cells. J. Immunol. 126:2011. Laurell, C. B. and J. O. Jeppson. 1975. Protease inhibitors in plasma. In The Plasma Proteins. Edited by F. W. Putnam. Academic Press, New York. 229. Lust, J. A., V. Kumar, R. C. Burton, S. P. Bartlett, and M. Bennet. 1981. Heterogeneity of natural killer cells in the mouse. J. Exp. Med. 154:306. MacDonald, H. R. and G. D. Bonnard. 1975. Comparison of effector cells involved in cell-mediated lympholysis and antibody dependent cell-mediated cytotoxicity in man. Scand. J. Immunol. 4:129. Martz, E., W. L. Parker, M. K. Gately, and C. D. Tsoukas. 1983. The role of calcium in the lethal hit of T lymphocyte-mediated cytolysis. In Advances in Experimental Medicine and Biology, Vol. 146. Edited by W. R. Clark and P. Goldstein. Plenum Press. 121. McCoy, J. L., R. B. Herberman, E. B. Rosenberg, F. C. Donnelly, P. H. Levine, and C. Alford. 1973. Natl. Cancer Inst. Monogr. 37:59. Meruelo, D., A. Paolino, N. Flieger, and M. Offer. 1980. Definition of a new T lymphocyte cell surface antigen, Ly 11.2. J. Immunol. 125:2713. Michaelson, J., M. Scheid, and E. A. Boyse. 1979. Biochemical features of Ly-5 alloantigen. Immunogen. 3:193. Minato, N., L. Reid, H. Cantor, P. Lengyel, and B. R. Bloom. 1980. Mode of regulation of natural killer cell activity by interferon. J. Exp. Med. 152:124. Minato, N., L. Reid, and B. R. Bloom. 1981. On the 39 heterogeneity of murine natural killer cells. J. Exp. Med. 154:750. Minato, N. and B. R. Bloom. 1982. Heterogeneity and regulation of the natural killer cell system. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 131. Morishima, Y., S. Ogata, N. H. Collins, B. Dupont, and K. 0. Lloyd. 1983. Carbohydrate differences in human high molecular weight antigens of B— and T-cell lines. Immunogen. 15:529. Murphy, J. W., and D. O. McDaniel. 1982. In vitro effects of natural killer (NK) cells on Cryptococcus neoformans. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 1105. Nabel, G., L. R. Bucalo, J. Allard, H. Wigzell, and H. Cantor. 1981a. Multiple activities of a cloned cell line mediating natural killer cell function. J. Exp. Med. 153:1582. Nabel, G., M. Fresno, A. Chessman, and H. Cantor. 1981b. Use of cloned populations of mouse lymphocytes to analyze cellular differentiation. Cell. 23:19. Nunn, M., J. Djeu, D. Lavrin, and R. B. Herberman. 1973. Natural cytotoxic reactivity of rat lymphocytes against syngeneic Gross leukemia. Proc. Amer. Assoc. Cancer Res. 14:87. Okumura, K., S. Habu, and K. Shimamura. 1982. The role of asialo GM1+ (GA1+) cells in the resistance to transplants of bone marrow or other tissues. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 1527. Ojo, B., and H. Wigzell. 1978. Natural killer cells may be the only cells in normal mouse lymphoid cell populations endowed with cytolytic ability for antibody-coated tumor target cells. Scand. J. Immunol. 7:297. Oldham,R. K., and R. B. Herberman. 1973. Evaluation of cell-mediated cytotoxic reactivity against tumor associated antigens with 125-I-iododeoxyuridine labeled target cells. J. Immunol. 111:1862. Omary, M. B., I. S. Trowbridge, and H. A. Battifora. 1980. Human homologue of murine T200 glycoprotein. J. Exp. Med. 152:842. 40 Omary, M. B., I. S. Trowbridge, and M. P. Scheid. 1980. T200 cell surface glycoprotein of the mouse: polymorphism defined by the Ly-5 system of alloantigens. J. Exp. Med. 151:1311. Ortaldo, J. R., H. D. Kay, and G. D. Bonnard. 1977. In vitro induction of natural killer (NK) and killer (K) cells. In Regulatory Mechanisms of Lymphocyte Activation. Proceedings of the Eleventh Leukocyte Culture Conference. Edited by D. D. Lucas. Academic Press, New York. Ortaldo, J., R. Herberman, and S. Pestka. 1982. Augmentation of human natural killer cells with human leukocyte and human recombinant leukocyte interferon. In NK and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 349. Paige, C. J., E. F. Figarella, M. J. Cuttito, A. Cahan, and O. Stutman. 1978. Natural cytotoxic cells against solid tumors in mice. II. Some characteristics of the effector cells. J. Immunol. 121:1827. Perussia, B., G. Trinchieri, and J.-C. Cerottini. 1979. Analysis of human lymphocyte subpopulations responsible for antibody—dependent and spontaneous cell-mediated cytotoxicity. Transplant Proc. 11:793. Pollack, S. B., M. R. Tam, R. C. Nowinski, and S. L. Emmons. 1979. Presence of T cell-associated surface antigens on murine NK cells. J. Immunol. 123:1818. Reynolds, C. W., T. Timonen, and R. B. Herberman. 1981. Natural killer cell activity in the rat. I. Isolation and Characterization of the effector cells. J. Immunol. 127:282. Quan, P.-C., T. Ishizaka, and B. R. Bloom. 1983. Studies on the mechanism of NK cell lysis. J. Immunol. 128:1786. Roder, J. C. and R. Kiessling. 1978. Target-effector interaction in the natural killer cell system. I. Co-variance and genetic control of cytolytic and target-cell-binding subpopulations in the mouse. Scand. J. Immunol. 8:135. Roder, J. C., R. Kiessling, P. Biberfeld, and B. Anderson. 1978. Target-effector interaction in the natural 41 killer (NK) cell system. II. The isolation of NK cells and studies on the mechanism of killing. J. Immunol. 121:2509. Roder, J. C., A. Rosen, E. M. Fenyo, and F. A. Troy. 1979. Target—effector interaction in the natural killer cell system: isolation of target structures. Proc. Natl. Acad. Sci.(USA). 76:1405. Roder, J. C., S. Argov, M. Klein, C. Petersson, R. Kiessling, K. Andersson, and M. Hansson. 1980. Target-effector cell interaction in the natural killer cell system. V. Energy requirements, membrane integrity, and the possible involvement of lysosomal enzymes. Immunology 40:107. Rosenberg, E. B.,R. B. Herberman, P. H. Levine, R. H. Halterman, J. L. McCoy, and J. R. Wunderlich. 1973. Lymphocyte cytotoxicity reactions to leukemia-associated antigens in identical twins. Int. J. Cancer. 9:648. Rosenberg, E. B., J. L. McCoy, S. S. Green, F. C. Donnelly, D. F. Siwarski, P. H. Levine, and R. B. Herberman. 1974. Destruction of human lymphoid tissue-culture cell lines by human peripheral lymphocytes in 51-Chromium—release cellular cytotoxicity assays. J. Natl. Cancer Inst. 52:345. Saksela, E., O. Carpen, and I. Virtanen. 1982. Cellular secretion associated with human natural killer activity. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 983. Scheid, M. P., and D. Triglia. 1979. Further description of the Ly-5 system. Immunogenetics. 9:423. Seaman, W. E., T. D. Gindhart, J. S. Greenspan, M. A. Blackman, and N. Talal. 1979. Natural killer cells, bone, and the bone marrow: studies in estrogen-treated mice and congenital osteopetrotic (mi/mi) mice. J. Immunol. 122:2541. Seaman, W. E., N. Talal, L. A. Herzenberg, L. Herzenberg, and J. Ledbetter. 1981. Surface antigens on mouse natural killer cells: use of monoclonal antibodies to inhibit or enrich cytotoxic activity. J. Immunol. 127:982. Seeley, J. K., and S. H. Golub. 1978. Studies on cytotoxicity generated in human mixed lymphocyte cultures. I. Time course and target spectrum of several distinct concomitant cytotoxic activities. 42 J. Immunol. 120:1415. Seeley, J. K., G. Masucci, A. Poros, E. Klein, and S. H. Golub. 1979. Studies on cytotoxicity generated in human mixed lymphocyte cultures II. Anti-K562 effectors are distinct from allospecific CTL and can be generated from NK-depleted T cells. J. Immunol. 123:1303. Siadak, A. W., and R. C. Nowinski. 1980. Identification of Ly—5 and T200 antigens on identical cell surface proteins. J. Immunol. 125:1400. Sparrow, R. L. and I. F. McKenzie. 1983. A function for human T200 in natural killer cytolysis. Transplantation. 36:166. Springer, T., G. Galfre, D. S. Secher, and C. Milstein. 1978. Monoclonal xenogeneic antibodies to murine cell surface antigens: identification of novel leukocyte differentiation antigens. Eur. J. Immunol. 8:539. Stutman, 0., C. J. Paige, and E. Figarella. 1978. Natural cytotoxic cells against solid tumors in mice. I. Strain and age distribution and target cell susceptibility. J. Immunol. 121:1819. Stutman, O., P. Dien, E. Wisun, and E. C. Lattime. 1980. Natural cytotoxic cells against solid tumors in mice: blocking of activity by D-mannose. Proc. Natl. Acad. Sci.(USA). 77:2895. Stutman, 0., E. C. Lattime, and E. F. Figarella. 1981. Natural cytotoxic cells against solid tumors in mice: a comparison with natural killer cells. Federation Proceedings, 40:2699. Tagliabue, A., D. Boraschi, S. Alberti, and W. Luini. 1982. Large granular lymphocytes as effector cells of natural killer activity in the mouse. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 25. Takasugi, M., M. R. Mickey, and P. I. Terasaki. 1973. Reactivity of lymphocytes from normal persons on cultured tumor cells. Cancer Research. 33:2898. Takasugi, M., D. Koide, and A. Ramseyer. 1977. Specificities in natural cell mediated cytotoxicity by the cross-competition assay. Int. J. Cancer. 19:291. Talmadge, J. E., K. M. Meyers, D. J. Prieur, and J. R. 43 Starkey. 1980. Role of natural killer cells in tumor growth and metastasis: C57BL/6 normal and beige mice. J. Natl. Cancer Inst. 65:929. Targan, S. R., and W. Newman. 1983. Definition of a ‘trigger' stage in the NK cytolytic reaction sequence by a monoclonal antibody to the glycoprotein T—200. J. Immunol. 131:1149. Targan, S., and N. Stebbing. 1982. In vitro interactions of purified cloned human interferons on NK cells: enhanced activation. J. Immunol. 129:934. Tartakoff, A. M. 1981. The golgi complex: crossroads for vesicular traffic. Int. Rev. Exp. Pathol. 22:227. Timonen, T., E. Saksela, A. Rankl, and P. Hayry. 1979. Fractionation, morphological and functional characterization of effector cells responsible for human natural killer activity against cell-line targets. Cell. Immunol. 48:133. Timonen, T., E. Saksela, I. Virtanen, and K. Cantell. 1980. Natural killer cells are responsible for the interferon production induced in human lymphocytes by tumor cell contact. Eur. J. Immunol. 10:422. Timonen, T., J. R. Ortaldo, and R. B. Herberman. 1982. Analysis by single cell cytotoxicity assay of natural killer (NK) cell frequencies among human large granular lymphocytes and of the effects of interferon on their activity. J. Immunol. 128:2514. Timonen. T. 1983. Characteristics of fresh and cultured natural killer cells. Clinics Immunol. Allergy, 3:465. Toth, M. K., and G. A. Granger. 1979. The human lymphotoxin system. VI. Identification of various saccharides on LT molecules and their contribution to cytotoxicity and charge heterogeneity. Mol. Immunol. 16:671. Trinchieri, G., D. Granato, and B. Perussia. 1981. Interferon—induced resistance of fibroblasts to cytolysis mediated by natural killer cells: specificity and mechanism. J. Immunol. 126:335. Trowbridge, I. S., P. Ralph, and M. J. Bevan. 1975. Differences in the surface proteins of mouse B and T cells. Proc. Natl. Acad. Sci. (USA). 72:157. Trowbridge, I. S., and C. Masauskas. 1976. Immunological 44 properties of murine thymus—dependent lymphocyte surface glycoproteins. Eur. J. Immunol. 6:557. Trowbridge, I. S., M. Nilsen-Hamilton, R. T. Hamilton, and M. J. Bevan. 1977. Preliminary characterization of two thymus—dependent xenoantigens from mouse lymphocytes. Biochem. J. 163:211. Trowbridge, I. S. 1978. Interspecies spleen—myeloma hybrid producing monoclonal antibodies against mouse lymphocyte surface glycoprotein, T200. J. Exp. Med. 148:313. Tung, J.—S., M. P. Scheid, M. A. Pierotti, U. Hammerling, and E. A. Boyse. 1981. Structural features and selective expression of three Ly-5+ cell—surface molecules. Immunogenetics. 14:101. Tung, J.—S., M. P. Scheid, and M. A. Palladino. 1983. Different forms of Ly—5 within the T-cell lineage. Immunogenetics. 17:649. Ullberg, M., and M. Jondal. 1981. Recycling and target binding capacity of human natural killer cells. J. Exp. Med. 153:615. van Eijk, V. W., and P. F. Muhlradt. 1981. Carbohydrate- labeled glycoproteins as markers of human lymphocyte subsets stimulated with mitogen and alloantigen. Eur. J. Biochem. 115:23. Warner, J. F., and G. Dennert. 1982. Effects of a cloned cell line with NK activity on bone marrow transplants, tumor development and metastasis in vivo. Nature. 300:31. Watson, A., B. Dunlap, and F. H. Bach. 1981. The biosynthesis of Ly—S in T and B cells. J. Immunol. 127:38. Welsh, R. M., R. M. Zinkernagel, and L. A. Hallenbeck. 1979. Cytotoxic cells induced during lymphocytic choriomeningitis virus infection in mice. II. ‘Specificities' of the natural killer cells. J. Immunol. 122:475. Welsh, R. M., K. Karre, M. Hansson, L. A. Kunkel, and R. W. Kiessling. 1981. Interferon—mediated protection of normal and tumor target cells against lysis by mouse natural killer cells. J. Immunol. 126:219. Woody, J. N., M. Feldmann, P. C. L. Beverly, and I. F. C. McKenzie. 1977. Expression of alloantigens Ly—5 and Ly-6 on cytotoxic effector cells. J. Immunol. Wright, Young, 45 118:1739. S. C. and B. Bonavida. 1981. Selective lysis of NK—sensitive target cells by a soluble mediator released from murine spleen cells and human peripheral blood lymphocytes. J. Immunol. 126:1516. W. W., S. I. Hakamori, J. M. Durdik, and C. S. Henney. 1980. Identification of ganglio-N—tetraosylceramide as a new cell surface marker for murine natural killer (NK) cells. J. Immunol. 124:199. Zarling, J. M., and F. H. Bach. 1982. Stimulation of lymphocytes with allogeneic normal cells and autologous lymphoblastoid cell lines: distinction between NK—like cells and cytotoxic T lymphocytes by monoclonal antibodies. In NK Cells and Other Natural Effector Cells. Edited by R. B. Herberman. Academic Press, New York. 791. SURFACE Ly-S GLYCOPROTEIN IN MURINE NATURAL KILLER (NK) CELL I‘ DEVELOPMENT, TARGET BINDING AND CYTOTOXICITY MICHAEL H. ZAROUKIANi WALTER J. ESSELMAN and HAROLD C. MILLER From the Department of Microbiology and Public Health and the Department of Medicine, College of Human Medicine, Michigan State University, East Lansing, Michigan 48824—1101 46 47 Footnotes: *Recipient of Physician's Research Training Fellowship PRTF-10 from the American Cancer Society. a Supported by U.S. Public Health Service grants CA-13396 and CA-24437 from the National Cancer Institutes, National Institutes of Health. This is article No. 11322 from the Michigan Agricultural Experiment Station. 1Abbreviations: avidin-FITC, fluoresceinated avidin; B6, C57BL/6 mice; biotin/avidin-FITC, sequential reaction with biotinylated anti- mouse IgG and fluoresceinated avidin; C, complement; CM, conditioned medium; ConA, concanavalin A; CTL, cytotoxic T lymphocytes; FBS, fetal bovine serum; GPS, guinea pig serum; HEPES, N-2—hydroxyethylpiperazine- N'-2-ethanesulfonic acid; mAb, monoclonal antibody; NK, natural killer; PBS, phosphate buffered saline; sIg, surface immunoglobulin; TBC, tumor binding cells. ABSTRACT The role of surface Ly-S glycoprotein expression in the binding and lysis of susceptible tumor targets by natural killer cells was studied using nonphagocytic, nylon wool nonadherent, T cell-depleted splenocytes from 6-8 week old C57BL/6 mice which were reacted with anti-Ly-S serum in the presence or absence of a source of complement. A conjugate assay was used to demonstrate that abrogation of tumor cell lysis by anti-Ly—S serum involved the inhibition of NK cell binding to susceptible YAC-1 targets. Indirect immunofluorescent labeling and cell sorting were used in the physical separation of Ly-S+ and Ly-5‘ cells to examine the effect of interferon and interleukin preparations on Ly-S expression and NK activity. Three hour treatment of sorted Ly—S‘ cells with murine G., B interferon resulted in conversion of 22% Of the cells to an Ly—5+ phenotype, as well as a significant increase in the percent specific lysis of NK-susceptible YAC-1 targets when compared to freshly sorted Ly-S- cells (29.5 1 1.9 vs 2.6 1 4.0; p €001). £3 23339 proliferation of sorted Ly-S- cells was induced by three week culture in 40% super- natant from ConA stimulated BALB/c splenocytes (CM), followed by repeat analysis of Ly-S expression and cytotoxic activity. Cell sorter purified Ly-S— cells cultured in CM acquired substantial surface Ly-5 with concomitant high levels of NK activity that remained susceptible to inhibition by anti—Ly—5 serum. The data presented suggest that surface Ly-S glycoprotein expression is important for binding of NK cells to YAC-1 targets. In addition, Ly-S' precursors of NK cells are present in murine splenic tissues and can be induced by CM to become highly active NK cells with increased surface Ly-S expression. The persistent susceptibility of these cells to inhibition of cytotoxic activity by 48 49 anti-Ly-S serum provides additional evidence of an important role for the Ly-S glycoprotein ill the natural killer cell cytolytic mechanism against certain targets. Introduction The Ly-S glycoprotein is a transmembrane alloantigen expressed on the surface of most cells of lymphopoietic or myelopoietic origin in mice (1,2) and appears to be a homologue of the leukocyte common antigen (LCA) in humans (3,4). This cell surface glycoprotein has been identi- fied on cytotoxic T lymphocytes (CTL) and natural killer (NK) cells (5,6), yet only NK cell-mediated cytotoxicity was shown to be inhibited by treatment with specific antiserum in the absence of complement (7,8). This property indicated that fundamental differences exist in the two forms of cytotoxicity and that the Ly—S glycoprotein may function in one or more steps of the NK cell lytic mechanism (i.e., binding, programming and killer cell-independent lysis) against certain targets. Addition— ally, the consistent expression of the Ly—5 glycoprotein on functionally mature NK cells and its apparent importance in the lytic process make more likely the possibility that cytolytically inactive NK precursor cells exist which lack significant surface Ly-5 glycoprotein expression (Ly-5'). The isolation and purification of viable Ly-5_ cells allows for the analysis of regulatory factors that induce the proliferation and functional maturation of pre-NK cells as well as determining the relationship between Ly—S expression and cytotoxic capability. Of the available methods for enriching lymphoid cell suspensions in Ly-S- cells, the use of anti-Ly-S serum and complement (9) appears to be unsatisfactory since Ly-5 expression correlates poorly with complement- 50 mediated cytotoxic susceptibility and T cells are selectively eliminated by this procedure (10). Viable NK cells could persist in such prepara- tions but not be identifiable in a standard NK assay due to the inhibi- tory effect of anti-Ly-S serum. Similarly, recovery of NK activity in such preparations upon treatment with interferon or interleukins could be due either to recruitment of pre-NK cells or recovery of antibody- bound NK cells or both. Using 12 yitgg culture and functional analysis of cell sorter puri- fied Ly-S- and Ly-S+ murine splenocytes, we report here that Ly—S' pre- cursors of NK cells are present and can be induced by interferon and interleukin preparations to become highly active NK effector cells. Cytotoxic activity is accompanied by increased surface expression of Ly— 5 glycoprotein. The ability of anti-Ly-S to inhibit cytotoxicity suggests that the Ly-S glycoprotein participates in the cytolytic process. We further report that the Ly-5 glycoprotein is a mediator of murine natural killer cell binding to YAC-1 tumor cells. Materials and Methods 111363. C57BL/6J, BALB/cJ, A.SW/SnJ and SJL/J mice were purchased from the Jackson ‘Laboratory' (Bar Harbor, ME) and maintained in our facility. Some of the C57BL/6J mice were also used as breeders. Animals used in these experiments were five to eight weeks of age. Target Cells. The NK-susceptible YAC-1 lymphoma of A/Sn mice and the NK-resistant P815 mastocytoma of DEA/2 mice were generous gifts from Dr. G. Cudkowicz, and were grown in RPMI—16’40 culture medium (Gibco Laboratories, Grand Island, NY), supplemented with 15 mM HEPES (Sigma Chemical Co., St. Louis, MO), penicillin/streptomycin (Gibco) and 10% heat-inactivated (56°C x 30 min) fetal bovine serum (Microbiological 51 Associates, Walkersville, MD). Cells were maintained in a 37°C, 5% C02, humidified incubator. Preparation of NK Effector Cells. C57BL/6J spleen cell suspensions were aseptically prepared in RPMI-1640 medium with 10% fetal bovine serum (FBS) as described previously ( 11). Lysis of erythrocytes was achieved by 10 second exposure of spleen cell suspensions to double distilled water followed by a return to isotonicity with concentrated phosphate buffered saline (PBS). Adherent and phagocytic cells were removed by 115 minute incubation with pre-washed carbonyl iron (50 mg/1 x 108 cells, 37°C, 5% C02) followed by application of a bar magnet to the side of the tube and aspiration of nonphagocytic cells. Alternatively, adherence to plastic surfaces (75 cm2 polystyrene tissue culture flasks, Corning Glass Works, Corning, NY) was employed and non- adherent cells collected. Nonadherent, nonphagocytic cells were depleted of B cells by passage through nylon wool columns (Fenwal Laboratories, Deerfield, IL) containing 0.6 gm nylon fiber loosely packed into 10 ml plastic syringes (Becton, Dickinson and Co., Rutherford, NJ) that were pre—incubated with medium and FBS. Four m1 7 7 containing 3 x 10 -5 x 10 cells were loaded into the columns and incubated 60 min in a 37°C, 5% C0 environment, after which nylon wool 2 nonadherent cells were collected from the bottom of the column by the dropwise addition of 25 ml medium and PBS to the top. Nylon wool nonadherent splenocytes were treated with monoclonal anti-Thy 1.2 (New England Nuclear, Boston, MA), and complement (guinea pig serum, GPS, Gibco) for 60 min at 37°C, 5% C0 to eliminate T cells. Cellular 2’ viability was determined by trypan blue exclusion. For some experi- ments, nylon wool column separation and anti-Thy 1.2 + GPS treatment was 52 replaced by discontinuous density gradient centrifugation of splenocytes in Percoll (Sigma), as described previously (12). Briefly, 100% Percoll (nine parts Percoll, one part 10X PBS) was diluted in medium to make final concentrations of 30, 40, 50, 60 and 70% (v/v), and gently layered in order of decreasing density in conical 15 ml plastic tubes. 1 x 108 plastic nonadherent cells were layered on top and centrifuged at 375 x G for 30 minutes at room temperature. By this procedure, NK effectors were concentrated in a band at the interface between the 50% and 60% Percoll layers. Cells were collected, washed in medium and kept on ice until use. Natural Killer Cell Cytotoxicity Assay. YAC-1 and P815 tumor cells 51 were labeled. with Cr, 'washed three times and placed into separate soft, V-shaped plastic microtiter wells at 4 x 10h cells/well. Specified numbers of NK cell-enriched spleen cells or sorted and _i_n_ 319:9 cultured splenocytes were added to the microtiter wells and the plates centrifuged at 100 rpm x 60 seconds. Certain wells were pre- treated with monospecific or monoclonal (New' England. Nuclear) anti- Ly 5.1, at dilutions of 1:10 and 1:1000, respectively. After incubation for 11-6 hours on a rotating platform (40 rpm) at 37°C, 5% C0 the 2’ plates were centrifuged at 400 rpm x 10 min and half of the supernatant was removed and counted in a gamma counter. The wells were individually cut and placed into tubes containing their respective supernatants and 51 total counts measured. Percent Cr release and specific lysis were calculated by the following formulas: % 51 2 x i supernatant cpm total cpm x 100% Cr release = 53 % release in test - % spontaneous release % maximal release - % spontaneous release x 100% % specific lysis = where the addition of medium and FBS was used to determine spontaneous release and 3% perchloric acid was used to achieve maximal release. Target Cell Binding Assay. 1 x 105 NK effector cells in plastic round bottom tubes were mixed with an equal number of YAC-1 or P815 target cells in 0.2 ml culture medium. Cell mixtures were incubated at 37°C x 8 minutes, centrifuged at 200 x G, 5 min and reacted for 30 min on ice. Cells were gently aspirated using a pasteur pipette and examined microscopically. The percentage of single effector (small) cells bound to single target (large) cells was determined visually after counting 100 effector cells. Each sample was performed in triplicate. Fluorescent LabelingL Cytofluorimetric Analysis and Cell Sorting. NK ‘effector cells were adjusted to 1.2 x 107 cells/ml in PBS and incubated for 30 minutes with an equal volume of murine monospecific or monoclonal anti-Ly-S at final dilutions of 7 1:10 and 1:1000, respectively. Cells were washed, centrifuged and reacted sequentially on ice with biotinylated anti-mouse IgG and fluoresceinated avidin (avidin-FITC, Vector Laboratories, Burlingame, CA), at dilutions of 1:80 and 1:100, respectively. Cells treated with fluoresceinated avidin with and without biotinylated anti-mouse IgG served as controls for the determination of specific fluorescence. Cells were analyzed cytofluori- metrically in an Orthocytofluorograf 50H cell sorter/2150 comuputer system, using narrow angle helium-neon laser scatter and right angle argon laser fluorescence parameters. Cells were sorted into Ly—5+ and Ly-S- pOpulations and collected aseptically into RPMI-16110 medium and 50% FBS. Sorted cells were washed, centrifuged and resuspended in 54 culture medium for subsequent treatment and analysis. Viability of sorted cells as determined by trypan blue exclusion always exceeded 80%. Interferon Treatment and Analysis of Sorted Ly-5- Splenocytes. Sorted Ly-S' cells were incubated 3 hours with 5000 units of murine interferon (Lee Biomolecular, San Diego, CA), in culture medium or in medium alone at 37°C. Cells were then assessed cytofluorimetrically for acquired expression of Ly-5 using anti-Ly-5 IgG, biotinylated anti-mouse IgG and avidin-FITC as described above. A comparison of the natural killer cell activities of unfractionated and sorted Ly5+ and LyS- C57BL/6 splenocytes was made after incubation of effector cells in the . 1 presence or absence of murine Interferon. The 5 3 Cr release assay was performed as outlined above, except that 5 x 10 target cells/well were used, with an effector:target ratio of 12.5. In Vitro Culture of Sorted Ly-S- and Ly-5+ Splenocytes. Conditioned medium (CM) was used to support the i2 Agar—o culture of sorted Ly-5- and Ly-5+ splenocytes and was generated by 48 hour culture of BALB/c spleen cells (5 x 106/ml) with 10 g/ml concanavalin A (ConA, Sigma) in culture medium. Residual ConA was removed by the addition of 1 g/100 ml Sephadex G-50 (Pharmacia, Piscataway, NJ). The supernatant was passed through 0.2 micron Nalgene filters and tested for its ability to support the ig,yi£§g proliferation of C57BL/6J spleen cells. The CM used in these experiments was composed of 40% ConA BALB/c splenocyte culture supernatant and 60% RPMI-1640 medium containing 10% FBS. Production of Anti-Ly-S Serum. Six-week old SJL/J female mice were injected intraperitoneally with 1 x 108 spleen cells and thymocytes from female A.SW/SnJ mice every 2 weeks (13). Animals were aseptically bled from the tail vein on days 7 and 10 after the fourth and subsequent 55 injections. Sera were pooled and absorbed with SJL/J spleen and thymic cells (one part packed SJL/J cells, four parts sera, v/v). Monoclonal anti-Ly-5.1 IgG was purchased from New England Nuclear (Lot FPB 217). 2a Statistical Analysis. Data were analyzed using the two-tailed Student's t test to determine the significance of differences between experimental groups. Results Inhibition of NK Activity by Anti-Ly—S Serum. Splenic NK effector cells were isolated from 5—8 week old C57BL/6J mice and incubated with monospecific anti-Ly-S serum in microtiter wells for 60 min at 37°C in the presence or absence of endogenous or exogenous sources of comple- ment. 51 Cr labeled YAC-1 target cells were then added and the incuba- tion was continued an additional 4 hours. The specific lysis of NK- susceptible YAC-1 target cells was significantly diminished by anti-Ly-S serum, regardless of the presence or absence of a source of complement (Table I). Heat inactivation of the serum prior to its use ruled out an endogenous murine source of complement in the abrogation of NK activity. Furthermore, the high cellular viability of each treatment group indicated that NK cells were relatively resistant to complement mediated lysis, in that splenic T cells were effectively lysed by anti—Ly-5 + C (10, personal observations). lphibition of Targgt Cell Bindigg by Anti—Ly-S Sgggm. To test the hypothesis that anti-Ly-S serum interferes with the target binding step of NK-mediated cytolysis, discontinuous density gradient centrifugation was used to obtain splenocyte populations enriched in or depleted of NK effectors. Anti-Ly—S serum, normal mouse serum or an equal volume Of medium was added to the cell suspensions for 30 min, followed by removal 56 TABLE I Effect of Anti-Ly-5 Serum on NK Cell-Mediated Cytotoxicity against YAC-1 Target Cellsa Effector cell treatmentb % Specific lysisc Average % reduction in lysis Medium alone 45.1 i 4.8 Complement (C) control 41.8 i 6.1 7.3 Anti-Ly-S serum alone 15.1 1 0.6 66.5”e Heat-inactivated anti-Ly—S 5.9 1 3.3 86.9* serumd Anti-Ly-S serum + C 5.6 1 2.2 87.6* Heat-inactivated anti-Ly-5 2.3 i 1.1 94.9fl serum + C aEffector:target ratio = 25. bCell viability in all treatment groups exceeded 95%’ as measured by trypan blue exclusion. (o ). cMean: S.D. of triplicate samples n-1 dSerum heated to 56°C x 30 min to inactivate endogenous complement before use. 3' p < .001, ** p < .0001. 57 of excess antibody and the addition of YAC-1 or P815 tumor cells in a target cell binding assay (Table II). Pretreatment of NK effectors with anti—Ly-S serum substantially blocked binding to NK-susceptible YAC-1 targets (decreasing the number of conjugates/100 lymphocytes from 17 to 5), while normal mouse serum as a source of immunoglobulin had no effect. The NK-resistant P815 target was not bound by effector cells despite the absence of antiserum. In a parallel experiment, treatment of YAC-1 tumor cells with anti-Ly-S serum prior to the target cell binding assay resulted in an increase in conjugate formation (35.7 _-t_- 7.4%), indicating that an interaction of antiserum with target cells was not the cause of decreased binding by NK effectors. Cytofluorimetric Analysis of Ly-5 Expression on NK Effector Cells and Separation of Ly-S- and Ly-5+ Cells by Cell Sorting. NK effector cells underwent indirect immunofluorescent labeling with murine monoclonal anti-Ly-5.1 IgG (anti-Ly-S mAb), biotinylated anti-mouse IgG and fluoresceinated avidin. Cells labeled with avidin-FITC alone or in combination with biotinylated anti-mouse IgG (i.e., biotin/avidin-FITC) served as controls to determine background fluorescence and contaminating surface immunoglobulin positive (sIg+) B cells, respectively. To verify adequate removal of splenic B cells, which are labeled by the indirect immunofluorescence technique independent of their surface Ly—5 expression, unfractionated and nylon wool nonadherent splenocytes were compared for percent specific fluorescence after biotin/avidin-FITC treatment (Figure 1). Nylon wool column separation was shown to remove more than 90% of the sIg+ cells, leaving only 3% nonspecific fluorescence in the anti-Ly-S mAb treated suspension. Figure 2 shows the gating procedure used to define and sort Ly-5- and 58 TABLE II Inhibitory Effect of Anti-Ly-S Serum on Target Cell Binding by NK Cells Conjugates per 100 lymphocytesa % Percoll NK cells + NK cells + Target cell fraction NK cells normal serum anti-Ly—S c *b YAC-1 60 17.0 1 3.6 17.3 1 11.0 5.7 11.5 P815 60 0.3 1 0.6 NDd ND YAC-1 706 11.3 11.5 2.0 11.0 3.0 11.7 P815 70 0.3 + 0.6 ND ND ). aValues represent mean 1 S.D.of triplicate samples blinded to observer (oh_ bp < .005. 1 c60% Percoll fraction (enriched in NK cells). dND = not done. e70% Percoll fraction (depleted of NK cells). U) .J _J Lu u.1o . . ELL 3.. of”??? PO “troke. at“? A.’ o < .~-- ~31; “ Ag 1:." ..-" 65M. 3‘: 37%;", , 10 100 FLUORESCENCE INTENSITY Fig. 1. Fluorescence histograms of unfractionated (UF) and nylon wool nonadherent (NWNA), C57BL/6J splenocytes labeled with biotinylated anti— mouse IgG and fluoresceinated avidin. Relative cell number is plotted on the ordinate and fluorescence intensity on the abscissa. The dashed line ( ----- ) signifies the 99th percentile for autofluorescence of unlabeled spleen cells. The percentage of total cells exceeding this level of fluorescence in each of the treated groups is listed beneath their respective histograms. Nylon wool column separation substantially eliminates nonspecific immunofluorescence that is unrelated to specific surface marker expression in this system. 6O FLUORESCENCE INTENSITY Fig. 2. Fluorescence histogram of viable NWNA, anti-Thy 1 + GPS treated CS7BL/6J splenocytes treated with anti-Ly-S mAb + biotin/avidin- FITC. The percentages shown indicate those cells having defined Ly-S; and Ly-5 phenotypes. The baseline region between the Ly-5 and Ly-5 cell populations indicates the gating procedure used by the cell sorter to discriminate fluorescent from non-fluorescent cells. 61 Ly-5+ cells. The vast majority of nonphagocytic, nylon wool non- adherent, T cell-depleted splenocytes expressed relatively high amounts of surface Ly~5 glycoprotein. Ly-5' cells were purified by cell sorting and cytofluorimetric analysis of sorted cells indicated successful separation, with 4% positive cells in the Ly-5— fraction (Figure 3A) and 5% negative cells in the Ly—5+ fraction (Figure 3B). Effect of Interferon Treatment on Sorted Ly-S‘ Cells. Sorted Ly—S- cells were incubated for 3 hours at 37°C, 5% CO in the presence of 2, murine interferon (a + B, 5000 units) or in medium alone. Cell suspensions were then examined cytofluorimetrically for increased surface expression of Ly-5 glycoprotein (Figure 3). When compared with sorted Ly-5- and Ly—5+ cells (Figure 3A, 3B) both the medium control and interferon treated cells showed slight increases in fluorescence when labeled with biotinylated anti-mouse IgG + avidin—FITC (Figure 3C, 3E). However, only the interferon treated cells demonstrated a significant increase in specific surface Ly-5 expression under the same conditions, with 22% of the cells converting to a Ly-5+ phenotype (Figure 3D, 3F). TO define the relationship between interferon—induced increases in surface Ly—5 expression and acquired NK activity among sorted Ly~5_ cells, unsorted and cell sorter purified whole B6 splenocytes were compared in their ability to lyse NK—susceptible YAC-1 targets (Table III). Ly-S- sorted cells lacked cytotoxic activity immediately after sorting (2.6 1 4.0% specific lysis), but when cultured in medium containing interferon under conditions shown to increase surface Ly-5 expression, they acquired substantial NK activity, even at a relatively low effector:target ratio of 12.5 (29.5 1 1.9% specific lysis; p (.001). The degree of activity exceeded that generated by the incubation of 62 Fig. 3. Two-dimensional cytograms (narrow angle scatter versus right angle fluorescence) of sorted Ly-5_ C57BL/6J splenocytes before and after 3 hour incubation with 5000 units murine interferon (on + B) or medium control. Each group was then labeled with anti-Ly-S mAb (1:1000) + biotin/avidin—FITC and compared with a biotin/avidin-FITC control. A. Sorted Ly-S- +cells prior to incubation with interferon or medium. B. Sorted Ly-5 ceI+Lls indicating the degree of fluorescence characteristic of Ly-5 cells. C. Sorted Ly-5 cells incubated with medium alone x 3 hr, biotin/avidin—FITC control. D. Sorted Ly-S— cells incubated with medium alone x 3 hr, labeled with anti-Ly—5 + biotin/avidin-FITC. E. Sorted Ly-5- cells incubated with interferon x 3 hr, biotin/avidin—FITC control. F. Sorted Ly-S— cells incubated with interferon x 3 hr, labeled with anti-Ly-S + biotin/avidin-FITC. Note that an additional 22% of the cells (43.6-21.2), have converted to an Ly-5 phenotype when treated with the interferon preparation. 63 497. IK vavvv 1 I ‘7 7V7 VVV :0 40 so so 199 x 29 40 so so 188 Bl run psnlvn vs can FLIKI 1 El rfifivevo'svo‘e'ojxée '2'9'0'0'6'0'8'01160 F- PHD REDIY) vs GRN FL(X) Fun RED(Y) vs GRN FLIX) C) 108" (D LL] .J (.9 E K (I i ' 2b ' 4b ' 6b ' 8b ' 169 1 ' 2b ' ab ' 6b ‘ 8b ' 169 Fun psnrv: vs GRN FLIX) run REDIY) vs sew FLIXI FLUORESCENCE INTENSITY > Figure 3 64 Table III Effect of Interferon Treatment on NK Activity of Sorted Ly-5+ and Ly-5- C57BL/6 Splenocytesa % Specific Lysis of YAC-1 Targetsb’c Medium x Medium + Interferon (5000 u/ml) Splenocytes 0 Hoursd 3 Hours x 3 Hours Whole spleen 1.7 1 6.4 9.0 1 0.3 15.7 1 3.8 Sorted Ly-b‘+ 11.11 5.9 8.5 11.9 11.3 1 3.8 Sorted Ly-5- 2.6 + 4.0 19.7 1 6.0 29.5 1 1.9!!6 aWhole B6 splenocytes labeled with anti-Ly-5.1 + biotinlyated anti-mouse IgG + avidin-FITC as described in Materials and Methods, sorted into fluorescent (Ly-5+) and nonfluorescent (Ly~5-) populations. 51 b4 hour Cr-release assay, 5 x 103 YAC-1 targets/well, effector:target ratio 12.5. ). cMean: S.D. of triplicate samples (Q'n_1 dAssay of cells immediately after sorting; sorted cell populations )95% pure for defined cell population. e*p < .001. 65 whole B6 splenocytes with interferon (15.7 1 3.8%), or of interferon treated Ly-5+ sorted cells from whole spleen SUSpensions (11.3 1 2.5%). Effect of In Vitro CM Culture of Sorted Ly-5- and Ly—5+ Cells on Surface Ly—5 Glycoprotein Expression and NK Activity. Sorted Ly-S' and Ly—5+ cells were placed into culture for a period of 21 days with CM to support the proliferation of cells necessary for subsequent cytofluori- metric analysis of Ly-5 expression and NK activity. In both the Ly-S- and Ly-5+ cell cultures, CM induced the conversion of almost all of the cells to an Ly-5+ phenotype (Figure 4), with an average level of fluorescence at least 10 times greater than freshly isolated NK effector cells (Figure 2). CM cultured Ly—5_ cells were heterogeneous in their Ly-5 expression, with two distinct levels of flugrescence identified (Figure 4A). Sorted Ly-5+ cells appeared more uniform in their high level of surface Ly-5 expression after culture in CM, with a single peak level of fluorescence (Figure 4B). To identify a possible association between the conversion of Ly-5- cells to Ly-5+ cells, the development of NK activity and the role of the Ly-5 glycoprotein in the cytotoxic process, sorted and CM cultured cells were used as effectors in a standard NK assay (Table IV). Anti-Ly-5 mAb was added to some Of the wells to examine its effect on cytotoxicity. An effector to target ratio of 2 was used, a ratio that is usually insufficient to allow the demonstration of‘ NK activity in a standard 4 hour 51 Cr release assay (14). After showing that spent media from the CM cultures were without cytotoxic activity, it was clear that both of the sorted cell populations that were cultured in CM developed potent and selective cytotoxic activity against the NK-susceptible YAC-1 target. In addition, the cytotoxic process was significantly diminished 66 CELLS FLUORESCENCE INTENSITY RELATIVE NUMBER OF CELLS .8 O 100 1000 FLUORESCENCE INTENSITY Fig. 4. Fluorescence histograms of sorted Ly-S- and Ly-5+ CS7BL/6' splenocytes after 3 week culture in conditioned medium (CM). Note that the relative fluorescence intensity scale is 10 times higher than for cells analyzed prior to culture in CM. Cells were labeled with anti- Ly-5 + biotin/avidin-FITC in a manner identical to that for cells studied prior to culture in CM. The dashed line ( ----- ) represents the 85th percentile for fluorescence intensity of the biotin/avidin-FITC controls. A. Sorted Ly-5- cells cultured in CM, showing conversion to a highly Ly-5+ phenotype that is heterogenous in nature, with two populations of +cells bearing different amounts of Ly-5 discernible. B. Sorted Ly-S cells cultured in CM, showing a relatively homogenous population of cells expressing high amounts of Ly-S. 67 TABLE IV Increased NK Activity of Sorted C57BL/6 Spleen Cells Cultured 3 Weeks .12 vitro in Conditioned Medium and Inhibition by Anti-Ly-S mAb Medium or cell Antibody ; Specific lysisa Population Treatment YAC-1 P815 Spent medium from CM cultured Ly-5- none -2.1 1 2.1 1.6 1 1.4 sorted cells Spent medium from CM cultured Ly-5+ none -1.6 1 0.9 -0.8 1 1.5 sorted cells none 85.7 1 2.5 28.0 1 2.4 Sorted Ly-5+ cells . *b antl-Ly-S 5301 i 2.5 31.3 + 201 none 84.6 1 5.2 22.7 1 3.2 Sorted Ly-5- cells i anti-Ly-S 63.0 1 1.1 10.9 1 1.4 aEffector to target ratio = 2. b a p < .001. 68 but not eliminated by the addition of anti-Ly-S mAb in the absence of complement. Some cytotoxic activity 'was also observed against the usually NK-resistant P815 target. This process was not affected by the addition of anti-Ly-5 mAb to sorted Ly-5+ CM cultured cells, but an inhibitory influence against sorted Ly-5- CM' cultured effectors was suggested. Discussion In these studies, we have investigated the hypothesis that functional maturation of natural killer cells is accompanied by increased surface expression of Ly-5 glycoprotein. We also have identi- fied a mechanism by which antiserum directed against Ly-5 glycoprotein inhibits NK-mediated cytolysis of YAC-1 tumor cells. From these data, we conclude that Ly-S- NK precursors exist in murine splenic tissues and can be induced by conditioned medium to proliferate and differentiate to become potent NK effector cells. As determined by cytofluorometric analysis, these cytotoxic effector cells also express substantially greater amounts of surface Ly-5 glycoprotein than freshly isolated NK cells. In addition, the surface Ly-5 glycoprotein expressed on sorted and CM cultured effectors appears to be important for cellular function, in that specific antiserum diminished cytotoxicity against the NK- susceptible YAC-1 target. The lack of complete abrogation of NK activity was likely due to the polyclonal nature and heterogeneity of the cultured cells, with certain effector subsets highly inhibited by anti-Ly-S serum while other subsets were unaffected (13). While steric hindrance could be invoked as a nonspecific means of inhibiting NK cell- mediated cytotoxicity, this seems unlikely since Thy-1 glycoprotein is also highly expressed on CM cultured NK effectors (15,16), yet anti- 69 Thy-1 mAb does not inhibit cytotoxicity (our unpublished observations). It is also unlikely that cytotoxicity was due to a form of nonspecific CTL activity, since lysis was inhibited by anti-Ly-5 serum, which has been shown to specifically inhibit lysis by NK cells but not CTL (7). This is also supported by the finding that CTL-susceptible P815 targets were comparatively resistant to lysis (Table IV). Previous studies of natural killer cell precursors in humans have generally shown that pre-NK cells bind to target cells as well as NK effectors (17,18), and that interferon augments NK activity without increasing the number of tumor binding cells (TBC; 19,20). Our findings in mice on the importance of Ly-5 glycoprotein in YAC-1 target cell binding and the ability of interferon to induce surface Ly-5 expression among Ly-5- cells suggest that nonbinding pre-NK cells are also present. Such cells may be difficult to detect in standard assays in which TBC are assessed individually, since Ly-S- pre—NK cells appear to comprise at most only a few percent of typical NK—enriched splenocyte prepara- tions. In one other study of pre-NK cells, Minato, et al. (9), used anti- Ly-5 serum and complement to detect and deplete Ly-5+ murine spleno- cytes. They observed that 30-40% of CBA/J and 7% of BALB/c nude mouse splenocytes were susceptible to lysis by anti-Ly-S + C, with remaining viable cells having no NK activity. Interferon treatment was associated with a return of NK activity and additional cells susceptible to lysis by anti-Ly-S + C. From these results, they concluded that Ly—S- pre-NK cells were present in murine spleen and were stimulated by interferon to convert ‘30 LII-5+ functional NK cells. However, using a second step antibody, Seaman, et al. (7), showed that anti-T200 mAb, which also 70 specifically binds to the Ly-5 glycoprotein (21), lysed 82% of C57BL/6 splenocytes, suggesting that additional Ly-5+ cells may have been present in CBA/J and BALB/c nude spleens that resisted complement- mediated lysis. Further evidence in support of this possibility was obtained by Dennert et a1. (10), who found that in the absence of a second step antibody, anti-T200 mAb lysed more than 95% of CS7BL/6 thymocytes, but only 40% of spleen cells, 20-30% Of bone marrow cells and 5-10% of BALB/c nude spleen cells. By analytical cytofluorimetry of nonerythrocytic cells, they found that 78% of splenocytes, 91% of thymo- cytes and 81% of bone marrow cells were Ly-5+, confirming that a large percentage of Ly-5+ cells exist in bone marrow that resist lysis with anti-Ly-S + C. In addition, using tests of B and T cell function they were able to establish that anti-T200 mAb is selectively cytotoxic for thymocytes and mature T cells. Hence treatment with anti-Ly-S + C followed by the finding of abrogation of NK activity cannot be taken as presumptive evidence that mature NK cells have been eliminated, since, 1) many viable Ly-5+ cells remain in treated spleen suspensions; ii) Ly-5+ NK cells resist lysis with anti-Ly-S + C (Table I) and iii) NK cells treated with anti-Ly-S are inhibited in their cytotoxic activity without being lysed or removed from the cell suspension. Restoration of NK activity under such conditions by the addition of interferon could result either from recruitment of Ly-S- pre-NK cells or regeneration of lytic function among Ly-5+ NK cells. The use of a cell sorter to obtain Ly-5- cells and examine their ability to convert to Ly-5+ NK effectors allows more precise assessment of this process. While some of the effects seen in CM cultured cells could have resulted from proliferation of the few contaminating Ly-5+ 71 cells in the sorted Ly-5- preparation; no reliable method exists for the removal of all contaminating cells. This is especially true in view of the resistance Of Ly-5+ NK cells to lysis by anti—Ly-S + C. The probable significance of contaminating Ly—5+ cells is lessened by’ our observation that interferon treatment induced significant conversion of Ly—5— cells to a Ly-5+ phenotype without inducing proliferation (Figure 3). The presence of Ly-S- NK precursors in murine spleen is further supported by the results listed in Table III, which show that the conversion to a Ly-5+ phenotype is accompained by the acquisition of substantial NK activity. The observation that anti-Ly-S serum blocks binding of NK cells to YAC-1 targets suggests that the Ly-5 glycoprotein is involved in the binding step of NK mediated cytolysis and is, therefore, a NK cell receptor. It is of current interest in this laboratory to determine whether Ly-S also participates in subsequent steps of the NK-mediated cytolytic mechanism, namely, programming for lysis and killer cell- independent lysis (22). In a detailed hypothetical model supported by some experimental evidence, Hiserodt et a1. (23), characterized the likely properties of the NK cell receptor as, 1) binding to the target cell in a Mg++ dependent fashion, ii) possessing an external site for Ca++ binding, iii) being susceptible to proteolytic cleavage yet retaining its ability to interact functionally with the target cell, and, iv) having the ability to focus secreted NK cytotoxic factors to generate a lytic complex capable of irreversibly damaging the target cell membrane. The identification of Ly-5 glycoprotein as a mediator of target cell binding will allow further testing of this hypothetical model. Targan and Newman (24), provided recent evidence in the human NK 72 system that the Ly-5/T200 glycoprotein may participate in the early part of the Ca++-dependent programming step of NK cell-mediated cytolysis in that 13.1 mAb, which recognizes an epitope of T200, blocks lysis if added prior to Ca++ and does so without altering conjugate formation. A different anti-human T200 monoclonal antibody, F2.5, was recently shown to reduce the capacity of human NK cells to bind to K562 target cells (25), in agreement with our results in the murine system. Together, these data strongly suggest an important and fundamental role for Ly-S/T200 in the natural killer cell system. Any attempt to ascribe biological significance to the Ly-5 glyco- protein must contend with the fact that its expression is not unique to NK cells. Ly-5 is present on most cells of myelopoietic and lympho- poietic origin (2). However, heterogeneity has been observed in the molecular weight of the glyCOprotein among T cells, B cells and mono- cytes (26,27) despite apparent identity in peptide maps (27). While the molecular weight of Ly-5 expressed on NK cells has not been established, the glycoprotein may be sufficiently different in its glycosylation to give it unique properties. The clinical implications of the data presented here relate to recent evidence of the importance of NK cells in the immunosurveillance of cancer (reviewed in 28) in graft rejection (29) and in graft-versus— host (GVH) disease (30). Understanding the mechanism of NK receptor interaction with target cells and its precise role in the lytic process should lead to advances in the clinical immunotherapy of certain neo- plasms. In addition, the isolation and purification of NK cells or their precursors followed by 13111239 CM culture to induce proliferation and functional augmentation may represent an autologous means of 73 improving host defense mechaniams against cancer (31). The finding of an antibody that blocks NK cell function without altering cell viability may provide an additional strategy for the prevention of graft rejection and GVH disease in transplant recipients. Acknowledgements: We are grateful to the late Dr. G. Cudkowicz for generously providing the YAC-1 and P815 target cells used in these experiments. We also wish to thank L. Foutch for her technical assistance. References 1. Komuro, K., K. Itakura, E. A. Boyse, and M. John. 1975. Ly-5: a new T-lymphocyte antigen system. Immunogenetics 1:452. 2. Scheid, M. P., and D. Triglia. 1979. Further description of the Ly-5 system. Immunogenetics 9:423. 3. Omary, M. B., I. S. Trowbridge, and H. A. Battifora. 1980. Human homologue of murine T200 glycoprotein. J. Exp. Med. 152:842. 4. Judd, W., C. A. Poodry, S. Broder, S. M. Friedman, L. Chess and J. L. Strominger. 1980. High molecular weight antigen present on human T cells. Proc. Natl. Acad. Sci. USA 77:6805. 5. Woody, J. N., M. Feldman, P. C. L. Beverley, and I. F. C. McKenzie. 1977. Expression of alloantigens Ly-5 and Ly-6 on cytotoxic effector cells. J. Immunol. 118:1739. 6. Kasai, M., J. C. Leclerc, F. W; Shen, and H. Cantor. 1979. Identification of Ly-5 on the surface of 'Natural Killer' cells in normal and athymic inbred mouse strains. Immunogenetics 8:153. 10. 11. 12. 13. 14. 74 Seaman, W. E., N. Talal, Le. A. Herzenberg, Ld. A. Herzenberg, and J. A. Ledbetter. 1981. Surface antigens on mouse natural killer cells: use of monoclonal antibodies to inhibit or to enrich cyto- toxic activity. J. Immunol. 127:982. Newman, W. 1982. Selective blockade of human natural killer cells by a monoclonal antibody. Proc. Natl. Acad. Sci. USA 79:3858. Minato, N., L. Reid, H. Cantor, P. Lengyel, and B. R. Bloom. 1980. Mode of regulation of natural killer cell activity by interferon. J. Exp. Med. 152:124. Dennert, G., R. Hyman, J. Lesley, and I. S. Trowbridge. 1980. Effects of cytotoxic monoclonal antibody specific fbr T200 glyco- protein on functional lymphoid cell populations. Cell. Immunol. 53:350. Esselman, W. J., and H. C. Miller. 1977. Modulation of B cell response by glycolipid released from antigen stimulated T cells. J. Immunol. 119:1994. Kurmick, J. T., L. Ostberg, M. Stegagno, A. K. Kimura, A. Orn, and 0. SjOberg. 1979. A rapid method for the separation of functional lymphoid cell populations of human and animal origin on PVP-silica (Percoll) density gradients. Scand. J. Immunol. 10:563. Pollack, S. B., M. R. Tam, R. C. Nowinski, and S. L. Emmons. 1979. Presence of T-cell associated surface antigens on murine NK cells. J. Immunol. 123:1818. Tam, M. R., S. L. Emmons, and S. B. Pollack. 1980. Analysis and enrichment of murine natural killer cells with the fluorescence- activated cell sorter. J. Immunol. 124:650. 15. 16. 17. 18. 19. 20. 21. 75 Brooks, C. G., K. Kuribayaski, S. Olabuenaga, M. Feng, and C. S. Henney. 1982. Characterization of cloned murine cell lines having high cytolytic activity against YAC-1 targets. 111 NK Cells and Other Natural Effector Cells. R. Herberman, editor. Academic Press, Inc. 851. Brooks, C. G., K. Kuribayaski, G. E. Sale, and C. S. Henney. 1982. Characterization of five cloned murine cell lines showing high cytolytic activity against YAC-1 cells. J. Immunol. 128:2326. Targen, S. R. 1981. The dual interaction of prostaglandin E2 (PGEZ) and interferon (IFN) on NK lytic activation: enhanced capacity of effector-target lytic interactions (recycling) and blockade of pre-NK cell recruitment. J. Immunol. 127:1424. Ullberg, M., J. Merrill, and M. Jondal. 1981. Interferon-induced NK augmentation in humans: an analysis of target recognition, effector cell recruitment and effector cell recycling. Scand. J. Immunol. 14:285. Targan, S., and F. Dorey. 1980. Interferon activation of "pre- spontaneous killer" (pre-SK) cells and alteration in kinetics of lysis of both "pre-SK" and active SK cells. J. Immunol. 124:2157. Silva, A., B. Bonavida, and S. Targan. 1980. Mode of action of interferon-mediated modulation of natural killer cytotoxic activity: recruitment of pre-NK cells and enhanced kinetics of lysis. J. Immunol. 125:479. Siadak, A. W., and R. C. Nowinski. 1980. Identification of Ly-5 and T200 antigens on identical cell surface proteins. J. Immunol. 125:1400. 22. 23. 24. 25. 26. 27. 28. 29. 76 Hiserodt, J. C., L. J. Britvan, and S. R. Targan. 1982. Characterization of the cytolytic reaction mechanism of the human natural killer (NK) lymphocyte: resolution into binding, programming, and killer cell-independent steps. J. Immunol. 129:1782. Hiserodt, J. C., L. Britvan, and S. R. Targan. 1983. Studies on the mechanism of the human natural killer cell lethal hit: analysis of the mechanism of protease inhibition of the lethal hit. J. Immunol. 131:2705. Targan, S. R., and W. Newman. 1983. Definition of a 'trigger' stage in the NK cytolytic reaction by a monoclonal antibody to the glycoprotein T200. J. Immunol. 131:1149. Sparrow, R. L., and I. F. McKenzie. 1983. A function for human T200 in natural killer cytolysis. Transplantation 36:166. Trowbridge, I. S., P. Ralph, and M. J. Bevan. 1975. Differences in the surface proteins of mouse B and T cells. Proc. Natl. Acad. Sci. USA 72:157. Tung, J.-S., M. P. Scheid, M. A. Pierotti, U. Hammerling, and E. A. Boyse. 1981. Structural features and selective expression of three Ly—5+ cell surface molecules. Immunogenetics 14:101. Herberman, R. B. 1981. Natural killer (NK) cells and their possible roles in resistance against disease. Clin. Immunol. Rev. 1:1. Kiessling, R., P. S. Hochman, 0. Haller, G. M. Shearer, H. Wigzell, and G. Cudkowicz. 1977. Evidence for a similar or common mechanism for natural killer cell activity and resistance to hemo- poietic grafts. Eur. J. Immunol. 7:655. 30. 31. 77 LOpez, C., D. Kirkpartick, M. Sorell, R. V. O'Reilly, and C. Ching. 1979. Association between pretransplant natural killer cell activity and graft versus host disease after stem-cell transplanta- tion. Lancet II:1103. Herberman, R. B. 1983. Natural killer cells. 12 The Biology of Immunologic Disease. F. J. Dixon and D. W. Fisher, editors. Sinauer Associates, Inc. Sunderland, Massachusetts. 75. PRELIMINARY BIOCHEMICAL CHARACTERIZATION OF SURFACE Ly-5 GLYCOPROTEIN ON MURINE NATURAL KILLER CELLS. Michael H. Zaroukian+, Walter J. Esselman and Harold C. Miller, Department of Microbiology and Public Health and Department of Medicine, College of Human Medicine, Michigan State University, East Lansing, Michigan 48824-1101 Running Head: Characterization Of NK Cell Ly-5 Glycoprotein Footnotes: +Recipient of Physician's Research Training Fellowship PRTF-10 from the American Cancer Society. ”Supported by U.S. Public Health Service grants CA—13396 and CA-24437 from the National Cancer Institute, National Institutes of Health. Journal Article No. of the Michigan Agricultural Experiment Station. 78 79 1Abbreviations: avidin-FITC, fluorescein conjugated avidin; B6, C57BL/6J; BSA, bovine serum albumin; C, complement; °C, degrees centi- grade; ConA, concanavalin A; HEPES, N-2-hydroxyethylpiperazine-N'-2- ethanesulfonic acid; IgG, immunoglobulin G; KCIL, killer cell- independent lysis; KD, kilodaltons; LGL, large granular lymphocytes; M, molar; mAb, monoclonal antibody; Mr: relative molecular mass; 2-ME, 2- mercaptoethanol; NK, natural killer; NKCF, natural killer cytotoxic factor; NP—40, Nonidet-P40; PBS, phosphate buffered saline; PMSF, phenylmethylsulfonylfluoride; SAC, Staphylococcus aureus cells; SDS, sodium dodecyl sulfate; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; TEMED, N,N,N',N'-tetramethylethylenediamine. 2Zaroukian, M. H., W. J. Esselman and H. C. Miller. 1984. Surface Ly-5 glycoprotein in murine natural killer (NK) cell development, target binding and cytotoxicity. Submitted for publication. Abstract We previously reported the ability of monoclonal anti-Ly—S IgG to block effector cell participation in the target binding step of murine natural killer (NK) cell-mediated cytotoxicity against YAC-I lymphoma. In order to gain further insight into the physical properties of Ly-5 glycoprotein as a candidate surface molecule for NK cell interaction with target cells, the present study provides a preliminary biochemical characterization of Ly-5 on the plasma membrane of murine NK cell- enriched splenocytes. Nonadherent, T cell-depleted C57BL/6J splenocyte suspensions were enriched for NK cells by discontinuous Percoll density gradient. centrifugation. Lactoperoxidase-catalyzed radioiodination (of surface proteins on these intact cells followed by immunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) revealed a relative molecular mass of 205,000 for Ly-5 glycoprotein, with an electrophoretic mobility indistinguishable from that of macro- phage-associated Ly-5. While trypsin treatment of viable effector cells did not diminish the amount of immunofluorescently detectable surface Ly-5, SDS-PAGE analysis of the radioiodinated glycoprotein revealed that it had been cleaved to a peptide with an approximate Mr of 125,000 which resisted further trypsinization. The 125 KD fragment remained closely associated with the plasma membrane and could not be released by treat- ment of viable trypsinized cells with 1 M NaCl, 1% 2-mercaptoethanol or simple sugars known to interfere with NK cell-mediated cytotoxicity, appearing in the soluble phase only after detergent treatment of cells. 80 81 Reducing conditions revealed no evidence that the trypsin generated Ly-S fragment existed as a disulfide-linked heterodimer. The apparent Mr of the immunoprecipitable fragments generated in trypsinized viable cells and NP-40 solubilized membrane proteins were identical, suggesting that the fragment does not contain the intracytoplasmic peptide sequence of th (0 intact Ly-5 glycoprotein. The forces holding the Ly-5 fragment to the plasma membrane after trypsinization remain obscure. Introduction Natural killer (NK) cells have been implicated in various aspects of host defense mechanisms against cancer (1), microbial infections (2), in graft rejection (3) and in the regulation of normal hematopoiesis (4). The multiple activities of these cytotoxic effectors has prompted numerous investigations into the mechanism of NK cell-mediated cytolysis (5-8) and molecules important in the lytic process. Such efforts have led to the recognition that NK cell-mediated cytolysis occurs in several + -dependent pro- steps: 1) Mg++-dependent target cell binding, ii) Ca+ gramming of the NK effector for target cell lysis, iii) stimulus secre- tion of NK cytotoxic factor (NKCF), iv) focusing of NKCF on the target cell as a part of the killer cell-independent lysis (KCIL) step, and v) colloid osmotic lysis of the target cell. Based on experiments examining the effect of proteases and protease inhibitors on NK activity, Hiserodt et al. (8), proposed a model for NK cell-mediated lysis in which the properties of a putative NK receptor 82 were defined. According the this model, a surface molecule on NK cells that participates in target cell lysis could be expected to have the following characteristics: 1) a transmembrane structure, with an external domain and a cytoplasmic component, ii) the ability to mediate binding of the NK cell to its target in a Mg++-dependent manner, iii) a binding site for Ca++, iv) susceptibility to cleavage of the external domain by trypsin-like proteases in a manner that otherwise preserves its structural integrity, allowing continued interaction with the target cell during subsequent stages of the lyticmechanism, v) NKCF binding sites, and vi) the ability to integrate into a membrane attack complex that particpiates in tumor cell lysis. Attempts to identify candidate surface molecules on NK cells whose expression is important in NK cell-mediated cytotoxicity against tumor cells have generally involved the use of antibodies that inhibit NK activity at the effector cell level. One group of surface molecules identified in this manner as important in NK cell—mediated cytotoxicity in both murine and human NK cell systems is the Ly-5/T200 glycoprotein complex. Several studies have shown that antibodies directed against Ly-5 and T200 glycoproteins block NK activity at the effector cell level, even in the absence of complement (9-11). Ly-5 and T200 have since been shown to be distinct antigenic determinants on identical cell surface proteins ( 12). While molecular weight heterogeneity has been Observed for Ly-5 among T cells, B cells and macrophages (13,14), the 83 molecular weight form of Ly-5 expressed on NK cells and its relationship to cells of known hematopoietic lineage has yet to be characterized. Using lactoperoxidase-catalyzed radioiodination, immunoprecipita- tion and SDS-polyacrylamide gel electrOphoretic analysis of cell surface proteins from NK cell-enriched murine splenocyte populations, we report here that NK cell-derived Ly-5 glycoprotein migrates on SDS-PAGE with a relative molecular mass (Mr) of 205,000. When compared to the migration pattern of Ly-5 glycoproteins from a mixture of splenic T cells, B cells and macrophages, NK cell-derived Ly-5 migrates in a manner indistinguishable from ‘that of' macrophages. We further report that trypsin treatment of radiolabeled NK cell-enriched suspensions generates a fragment with a Mr of 125,000 that resists further trypsinization and retains the ability to bind anti-Ly-5 mAb. Combined with recent obser- vations that Ly-5 mediates murine NK cell binding to YAC-1 target cells aux! is expressed in increased amounts during differentiation of pre-NK cells to NK effectors, the results reported here suggest an important and fundamental role for Ly-5 glycoprotein in the natural killer cell system. Materials and Methods Cells and cell culture. Six to eight week old female C57BL/6J mice were obtained from the Jackson Laboratory (Bar Harbor, ME) for use in these experiments and served as donors of splenocytes and thymocytes. Respective cell suspensions were aseptically prepared as described pre- 84 viously (15), in RPMI 1640 medium (Gibco Laboratories, Grand Island, NY) containing L-glutamine and supplemented with 15 mM HEPES (Sigma Chemical Co., St. Louis, MO), penicillin-streptomycin (100 u/ml and 100 ug/ml, respectively) and 10% heat-inactivated (56°C x 30 min) fetal bovine serum (Microbiological Associates, Walkersville, MD). In some experi- ments, splenocytes were prOpagated by lg 11139 culture for 1 week in 40% conditioned medium which was generated by 48 hour culture of BALB/c spleen cells (5 x 106 /ml) with 1011g/ml concanavalin A (ConA, Sigma) in culture medium as previously described (16). Splenic NK cell isolation. Splenocyte suspensions were depleted of adherent and phagocytic cells by 45 minute incubation (37°C, 5% C02) in 7 Corning 75 cm2 polystyrene tissue culture flasks (2 x 10 cells/flask in 10 ml; Corning Glass Works, Corning, NY). Nonadherent cells were adjusted to 1.2 x 107 /ml and incubated with an equal volume of mono- clonal anti-Thy-1.2 (1:500, New England Nuclear, Boston, MA) and comple— ment (guinea pig serum, 1:5; Gibco) for 60 min at 37°C, 5% C02, in order to eliminate: T cells. Cells were then adjusted to 1:x 108/ml and subjected to discontinuous density gradient centrifugation in Percoll (Sigma), as described ‘previously (F7). Briefly, 100% Percoll (nine parts Percoll, one part 10X PBS), was diluted in RPMI-1640 medium with- out FBS to make final concentrations of 30, 40, 50, 60 and 70% (v/v), and gently layered in order of decreasing density in conical 15 m1 plastic tubes. 1 x 108 plastic nonadherent, T cell-depleted splenocytes were layered on top and centrifuged at 375 x G for 30 minutes at room 85 temperature. By this procedure, NK effectors were concentrated in a band at the interface between the 50% and 60% Percoll layers. Cells were collected, washed in medium and kept on ice until use. Enrichment 51 of NK cell activity was verified in a standard 4 hour Cr release assay as previously described.2 Fluorescent labeling and cytofluorimetric analysis. In order to determine the effect of trypsin treatment on surface Ly-5 expression, NK cell-enriched B6 splenocyte suspensions underwent indirect immunofluorescent labeling as previously describedz. Briefly, nonphago- cytic, nylon wool nonadherent, anti-Thy-1 + C resistant splenocytes were adjusted to 1.2 x 107 cells/ml in PBS and reacted on ice for 30 min with an equal volume of monoclonal anti-Ly-5.1 (1:500 in PBS + 5% BSA; New England Nuclear, Boston, MA). Cells were washed, centrifuged and reacted sequentially on ice with biotinylated anti-mouse IgG and fluoresceinated avidin (avidin-FITC, Vector Laboratories, Burlingame, CA), at dilutions of 1:40 and 1:500, respectively. Some cells were treated only with biotinylated anti-mouse IgG and avidin-FITC 1x3 serve as controls for the determination of specific fluorescence. Cells were analyzed cytofluorimetrically in an Orthocytofluorograf 50H cell sorter using narrow angle helium-neon laser scatter and right angle argon laser fluorescence parameters. Radiolabeling of cell surface proteins. Lactoperoxidase-catalyzed radioiodination of intact viable cells was performed essentially as pre- viously described (18). Briefly, 1 x 108 cells were suspended in 15GJ l 86 of PBS to which 100 pl of lactoperoxidase (1 mg/ml in 0.2 M PBS, Sigma Chemical Co.) was added, along with 20 “I of 2 M P0” (pH 7.3), 2 mCi 1251 (Amersham Corporation, Arlington Heights, IL, carrier-free Na 16.3 mCi/ug iodine), and 20 pl of 0.06% H202. After 15 min at 30°C, the reaction was terminated by adding 1.5 m1 KI—PBS (5 mg/ml KI). Cells were centrifuged at 1000 rpm x 10 min, washed in KI~PBS, repelleted and either resuspended in PBS for trypsin treatment or saved for immediate detergent solubilization of membrane proteins. Trypsinization of cell surface protein . Following radio- iodination, some cells were adjusted to 2 x 107/ml in PBS and subjected to proteolytic treatment with trypsin (1 mg/ml in PBS, 25 (ll/107 cells; Worthington Biochemical Corp., Freehold, NJ) for up to 20 minutes at room temperature. The reaction was terminated by the addition of Trasylol (250 KIU/107 cells; Bayer Werk Elberfeld, Bayer, Germany) and PMSF (0.1 M in acetone, 5 (ll/107 cells; Sigma). Cells were centrifuged at 1000 rpm x 10 min followed by separation of supernatant and pellet. Detergent solubilization of radioiodinated cell surface pgoteins. Radioiodinated pelleted cells were suspended in 1 ml 1% NP-40 (Particle Data Laboratory, Elmhurst, IL) in PBS containing 100 KIU/ml Trasylol protease inhibitor (FBA Pharmaceuticals, New York, NY), 0.02 mg/ml ovo- mucoid trypsin inhibitor (Sigma), and 0.02% NaN3. After 30 min on ice, suspensions were placed in 3 ml conical glass centrifuge tubes and centrifuged at 0°C, 10,000 rpm x 20 min. The supernate containing NP-40 solubilized radiolabeled surface proteins was then added to 3 ml Nazsou‘ 87 saturated ethanol and kept at —20°C overnight. The cold ethanol precipitate was collected by centrifugation at 13,000 ‘rpm x 20 min and the supernatant discarded. The precipitate was solubilized in 50 pl 1% NP-40, sonicated and centrifuged (Airfuge, Beckman Instruments, Inc., Palo Alto, CA) at 100,000 G x10 min. A second extraction using an additional 25 pl NP-40 was performed and pooled with the first solubili- zation to improve yield. Samples were counted in a gamma counter and stored at -20°C for subsequent immunoprecipitation. Immunoprecipitation of Ly-5 glycoprotein. NP-40 solubilized proteins were added to the pellet from 100 pl of 10% Staphlacoccus aureus cells (Pansorbin, Calbiochem, La Jolla, CA), to remove any proteins capable of binding nonspecifically to SAC and immunoglobulins binding specifically to the protein A component. After 15 min on ice, the suspensions were airfuged at 15 psi x 10 min and the supernatants retained. To each supernatant was added 20 pl of monoclonal anti-Ly-5 (1:100 in PBS with 5% BSA, New England Nuclear, Boston, MA) and the mix- ture kept on ice for 30 minutes, then added to SAC, mixed and reacted on ice an additional 15 minutes. The antigen-antibody-SAC complexes were collected by centrifugation, washed, and the radioiodinated Ly-5 glyco- protein liberated from the complexes by reaction with 2.3% SDS (Bio-Rad) sample buffer containing 5% 2-mercaptoethanol (Sigma), 0.0625 M Tris-Cl (Sigma) and 10% glycerol for 30 minutes at room temperature. Super- natants were recovered after airfuging at 15 psi x 10 min and boiled at 88 100°C x 2 min. Radioactivity was determined in a gamma counter and samples saved for SDS-polyacrylamide gel electrOphoresis. Use of simple sugars1 salt solutions and disulfide inhibitors. In order to investigate possible mechanisms to explain the observation that Ly-5 glycoprotein moieties remained associated with the cell surface despite proteolytic treatment, intact viable trypsinized cells were reacted with agents known to disrupt cell surface molecular interaction or covalent peptide linkages involving disulfide bonds. Simple sugar inhibitors of NK cell activity were used individually at concentrations of 5 mM and included mannose-6-phOSphate, N-acetyl glucosamine, D- mannose and L-fucose. Radioiodinated cells were trypsinized in the presence of one of these sugars and Ly-5 immunoprecipitates from pellet and supernatant compared for their content of Ly-5 glycoprotein or an immunoprecipitable degradation product. Other aliquots of cells were treated with either 1 M NaCl or 0.1% 2-ME after trypsin treatment of cells to determine whether either of these conditions resulted in the release of immunoprecipitable material from the cell surface. Comparison of the apparent molecular weight of the Ly35 fragments generated by trypsinization of intact cells and NP-40 solubilized cell surface proteins. To explore the possibility that the Ly~5 fragment generated by trypsin treatment of intact cells contained the trans- membrane and intracytoplasmic domains of the Ly-5 glycoprotein, at least a portion of which has been shown to be sensitive to trypsinization (20), viable radioiodinated 60% Percoll-fractionated splenocytes were 89 trypsinized as described above either before or after solubilization of membrane proteins in 1% NP-40. Trasylol and PMSF were added to trypsinized intact cells to halt proteolysis before the addition of NP-40. An additional aliquot of cells underwent trypsinization followed by the addition of NP-40 but a delay in the addition of Trasylol and PMSF for an additional 20 minutes. Following cold ethanol precipitation and solubilization in 1% NP-40, Ly-5 immunoprecipitates were collected and analyzed by SDS-PAGE as described below. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Ly-5 immunoprecipitates and whole cell surface radiolabeled protein samples in SDS buffer were electrophoresed on 1.5 mm thick 3-10% polyacrylamide (Bio-Rad) linear gradient gels which contained 0.1% SDS (w/v), 5% glycerol, 25% running gel buffer (1.5 M Tris-Cl, pH 8.8, Sigma), 0.05% ammonium persulfate (Fisher Scientific, Fair Lawn, NJ) and 0.033% TEMED (Aldrich Chemical Co, Milwaukee, WI). A linear gradient generating apparatus (MRA, Clearwater, FL) was employed for this purpose. Gels were allowed to polymerize under a layer of overlay buffer (0.375 M Tris-Cl, pH 8.8, 0.1% SDS), for 4 hours, after which a 2.5% acrylamide stacking gel was applied to the top of the linear gradient gel and allowed to polymerize x 30 min. The stacking gel components differed from the running gel only in percent acrylamide and the substitution of 25% stacking gel buffer (0.5 M Tris-Cl, pH 6.8) for running gel buffer. Tank buffer consisted of 0.025 M Tris-Cl, 0.192 M glycine and 0.1% SDS (pH 8.3). Bromphenyl blue served as the dye front. A Hoefer vertical 90 slab gel apparatus (Hoefer Scientific Instruments, San Francisco, CA) was used for electrophoresis, which was performed under 40 mA constant current conditions until the dye front reached the distal end of the gel. ‘High :molecular weight protein standards (BioRad Laboratories, Richmond, CA) were included in several lanes to allow precise molecular weight comparisons. After electrophoresis, gels were stained overnight in Coomassie blue solution (0.125% Coomassie blue, 50% methanol, 10% acetic acid) followed by sequential destaining (50% methanol, 10% acetic acid and 5% methanol, 7% acetic acid solution, respectively) and drying x 60 min on a Hoefer slab gel dryer. Autoradiography. Dried gels were placed on clean glass plates and autoradiography performed at -70°C using Kodak XAR-X-ray film with Cronex intensifying screens (DuPont Inc., Wilmington, DE). Results Enrichment of NK cell activity by discontinuous density gradient centrifugation. As large granular lymphocytes (LGL) with a relatively large cytoplasmic-to-nuclear volume ratio (19), NK cells can be physi~ cally separated from other lymphocytes by centrifugation through a discontinuous density gradient of Percoll. In the present study, gradients were established in 10% increments from 30% to 70%. NK cell activity was found to be highest in those cells concentrated in a band at the interface between 50% and 60% Percoll (Table I). No significant NK activity was observed in the other bands. As a.result, cells from 91 Table I Enrichment. of“ Natural Killer Cell Activity by Percoll Discontinuous Density Gradient Centrifugation of Plastic-nonadherent,‘T Cell-depleted Splenocyte Suspensions. 1 Specific lysisa Percoll fractionb YAC-1 P815 controlc 4.3 1 0.7 1.3 1 0.5 50% 5.9 1 0.6 7.5 1 2.3 60% 44.1 1 1.3”d 2.0 1 1.6 70% 11.5 + 2.2 2.1 + 0.6 a‘Values represent mean 1 S.D. of triplicate samples. YAC-1 is an NK— sensitive lymphoma of A/Sn mice. P815 is a NK-resistant mastocytoma of DEA/2 mice. Effectorztarget ratio = 25. bIndicates the percent Percoll layer that prevented further sedimentation of the cell band collected at its upper interface. cCells tested for' NK activity prior to fractionation through Percoll gradients. d' p < .00l. 92 the 60% Percoll band were used for biochemical characterization of Ly-5 glycoprotein. Persistence of immunofluorescently detectable Ly-5 on NK cells after trypsinization. It has been established that antisera to Ly-5 glycoprotein inhibit NK cell-mediated cytotoxicity (9-11). In order to determine whether the loss of NK activity observed following the treat- ment of NK effectors with trypsin (5) is associated with the loss of immunologically detectable surface Ly-5, NK cell—enriched B6 splenocyte suspensions were trypsinized (1 mg/ml x 20 min), followed by indirect immunofluorescent labeling with murine monoclonal anti-Ly—S, biotinylated anti-mouse IgG and avidin-FITC. Controls consisted of non— trypsinized cells labeled as just described or using only the biotinylated antibody and avidin—FITC. Figure 1A shows the specific Ly-5 immunofluorescence profile of NK-enriched B6 splenocytes before trypsin proteolysis of surface proteins. . Trypsin treatment did not decrease the amount of immunofluorescently detectable surface Ly-5 glycoprotein on these effector cells (Figure 1B). The fluorescense was not the result of nonspecific adherence of biotinylated antibody to trypsinized cells in that cells treated only with biotinylated antibody and avidin-FITC were no more fluorescent than unlabeled cells (data not shown). Molecular weight of Ly-5 glycopyotein on NK cells. SDS-PAGE evi- dence of molecular weight heterogeneity has been observed for Ly-5 glycoproteins isolated from cells of different hematopoietic lineage 93 RELATIVE NUMBER OF CELLS. .9: z? ”'34 ”(5‘11 l 1 10 100 FLUORESCENCE INTENSITY a) B m0 C: 3.; fix"; . 0 . s” ' ..' 53 :~ °‘ "9 5m ' LIJ LIJ 7‘ 3 m m f. I. 2 1. 3 33‘9"?" I 1 ‘ ’ .‘ “ M I I I 1 10 1 O O FLUORESCENCE INTENSITY Figure 1. Fluorescence histograms of control and trypsinized 11K cell- enriched B6 splenocyte suspensions labeled with urine monoclonal anti- Ly—S, biotinylated anti-mouse IgG and fluoresceinated avidin. Relative cell number is plotted on the ordinate and fluorescence intensity on the abscissa. A. Specific fluorescence of untrypsinized splenocytes labeled as above in comparison to control cells labeled only with biotinylated anti-mouse IgG and avidin-FITC. 82% of cells labeled with anti-Ly-S mAb express levels of fluorescence greater than the 99th percentile for fluorescence of control cells. B. Specific Ly-5 immunofluorescence profile of the same cells following tretment with 1 mg/ml trypsin x 20 min at 25°C. No diminution in immunofluorescently detectable surface Ly-5 is observed following trypsinization. 94 (13,14), resulting in estimates of 185-200 KD for T cells, 205 KD for macrophages and, 220 KD for B cells. The relative rarity of NK cells in murine spleen does not allow molecular weight determination or characterization of Ly-5 on these effectors without the removal of other contaminating cells. In order to determine the apparent molecular weight of NK cell surface Ly-5 glycoprotein and compare it with that of cells of known lineage, macrophage and T cell—depleted 60% Percoll fractionated B6 splenocytes underwent lactoperoxidase-catalyzed radio— iodination of surface proteins followed by immunoprecipitation with monoclonal anti-Ly-5 and SDS-PAGE analysis. When compared to an immuno- precipitate of radioiodinated whole B6 splenocytes, Ly-5 glycoprotein derived from the NK cell-enriched Percoll fraction migrated as a broad band with an apparent Mr of approximately 205,000, essentially equiva- lent to that of macrOphages and intermediate between that characteristic of B cells and T cells (Figure 2). Effect of trypsinization on NK cell surface Ly-5 glycoprotein. Trypsin treatment of intact and detergent solubilized membrane proteins of murine lymphocytes has been shown to result in the generation of a peptide with an approximate Mr of 100,000 that is immunoprecipitable with antisera to the T200 antigenic determinant (20). In that cell sorter analysis of trypsinized NK cell-enriched splenocytes revealed no diminution in the amount of immunofluorescently detectable surface Ly-S glycoprotein (Figure 1), experiments were directed to determine whether NK cell surface Ly-5 is cleaved into a peptide of similar molecular A B C 220— IIO — 92 - c7— 45-— Figure 2. Autoradiographic analysis of SDS-polyacrylamide gel electro- phoresis of Ly-S in-unoprecipitates from lactoperoxidase-catalyzed sur- face radioiodination of unfractionated and MK cell-enriched splenocyte suspensions. Lanes A and C show the labeled species precipitated by anti-Ly-S mAb treatment of NP-40 solubilized whole spleen suspensions cultured 1 week .i_n vitro in the presence of 40% conditioned medium as described in Materials and Methods. Lane B shows the labeled species precipitated by anti-Ly-S mAb treatment of NP-40 solubilized 60% Percoll-fractionated B6 splenocytes. 96 weight and, if so, whether trypsinization releases significant amounts of the peptide from the cell surface into the medium. Percoll- fractionated NK-enriched splenocytes were radioiodinated and trypsinized as described in Materials and Methods. Ly-5 immunoprecipitates of pro- teins from viable pelleted cells and the trypsin supernatant were generated and electrophoresed (n1 SDS-polyacrylamide gels. Figure 3 shows that trypsinizatiOn of these cells generates a fragment of Ly-5 which migrates as a broad band on SDS-PAGE with an approximate Mr of 125,000. Comparison of immunoprecipitates from viable trypsinized cells and the trypsin supernatant indicates that essentially all of the 125 KB Ly-5 fragment remains associated with the plasma membrane. That the Ly-5 glycoprotein is highly susceptible to proteolytic cleavage to a 125 KD fragment (21), is further suggested by the observation that even without deliberate trypsinization, failure to include antiproteases during detergent solubilization of membranes results in the conversion of significant amounts of the Ly-5 glycoprotein to the 125 KD fragments (Figure 3, lane C). Once generated, the 125 KD fragment resists further degradation by trypsin, even with high concentrations and prolonged incubation (Figure 4, lane C). Inability of simple sugars, high salt or 2-mercaptoethanol to release Ly-5 fragment from trypsinized cells. It was possible that certain covalent or noncovalent interactions were responsible for the persistent association of the 125 KD Ly-5 fragment with the cell surface after trypsinization. The potential mechanisms investigated in the pre- 97 2.20, 116 92 67 ' Figure 3. Effect of trypsinization of surface radioiodinated 60% Percoll-fractionated B6 splenocytes. Autoradiograph of SDS—polyacryl- amide gel of Ly-5 immunoprecipitates under conditions of deliberate protease inhibitor addition, trypsin treatment or absence of either agent. Lane A shows the Ly-5 immunoprecipitate of the supernatant medium in which intact 60% Percoll-fractionated cells were trypsinized, showing a lack of release of detectable immunoprecipitable Ly-5 from the plasma membrane. Lane B is an Ly—5 immunOprecipitate of 60% Percoll- fractionated splenocytes solubilized in NP-40 in the presence of the protease inhibitors Trasylol and trypsin ovomucoid inhibitor. Lane C represents the immunoprecipitate pattern obtained when radioiodinated cell surface proteins from B6 thymocytes were solubilized in NP-40 without protease inhibitors. Partial digestion to an immunoprecipitable fragment with an approximate molecular weight of 125 KD is apparent. Lane D is from an Ly—5 immunoprecipitate of cell surface proteins on viable 60% Percoll fractionated cells treated with trypsin. The vast majority of the Ly-5/T200 glycoprotein is digested to a fragment with an apparent molecular weight of 125 KD that remains associated with the cell surface. 98 .ABCDEFGHIJKL :'- 220'. ‘ 7.. 116 ‘ . 92 — Figure 4. Inability of simple sugars to release Ly-S fragment from trypsinized cells. Autoradiograph of Ly-S immunoprecipitates from trypsinized 60% Percoll-fractionated cells treated with various simple sugars at a concentration of 5 mM x 30 min. Immunoprecipitates of treated pelleted cells and the respective supernatants were compared for their relative content of the Ly-5 fragment. Lane A represents immuno- precipitated Ly-5 from radiolabeled 60% Percoll—fractionated cells prior to trypsin treatment. Lane B shows the Ly-5 immunoprecipitate of whole B6 splenocytes cultured 1 week 13 vitro in 40% conditioned medium. Lanes C through L are Ly-5 immunoprecipitates of radioiodinated 60% Percoll fractionated cells subjected to trypsinization followed by simple sugar addition: C, PBS control, pellet; D, PBS control, super- natant; E, mannose-6—phosphate, pellet; F3 mannose-6-phosphate, super- natnant; G, N-acetyl-D-glucosamine, pellet; H, N-acetyl-D-glucosamine, supernatant; I, D-mannose, pellet; J, D-mannose, supernatant; K, L- fucose, pellet; L, L-fucose, supernatant. None of the simple sugars tested released significant amounts of the 125 KD fragment from the cell into the supernatant. 99 sent study included, i) binding of carbohydrate portions of the glyco- protein to lectin-like receptors on the cell surface, ii) ionic or hydrogen bonding forces disruptible by solutions of high ionic strength, and iii) interchain disulfide bonding of the 125 KD fragment to another portion of the originally intact. Ly-S glycoprotein. that, anchors the protein to the plasma membrane, i.e., conversion of an intrachain disulfide bond to an interchain one by proteolytic cleavage of the pep- tide between the cystine residues. To investigate these possibilities, aliquots of Percoll-fractionated trypsinized viable cells were reacted individually with several carbohydrates known to interfere with NK acti- vity (D-mannose, mannose-6-phosphate, L-fUcose and N-acetyl-D-glucos- amine), 1 M NaCl or 0.1% 2—mercaptoethanol. Cells were centrifuged and immunoprecipitates of pelleted cells and supernatants after treatment compared for relative content of the 125 KD Ly-S fragment (Figures “,5). With the possible exception of 2-ME treatment (Figure 5, lane G) no significant release of the 125 KD fragment was seen with any of these treatments. Since the ability of 2-ME to induce release of the fragment from the cell surface even at concentrations up to 1% was only slight and not generally reproducible (data not shown), isolation and SDS-PAGE analysis was performed under reducing and nonreducing conditions to directly examine whether the trypsinized Ly-S fragment on intact cells exists as a disulfide-linked heterodimer. Uhder nonreducing conditions (Figure 6, lane B), the Ly-S fragment displayed slightly greater mobility (n1 SDS-PAGE than. when electrOphoresis was performed in the 100 AB C D E F G H 220— 116— 92— 67- A 45— 3? I Figure 5. SDS-polyacrylamide gel autoradiograph of Ly-S imunoprecipi- tates from trypsinized 60$ Percoll-fractionated splenocytes and trypsin supernatants. Trypsinized viable cells were pelleted and resuspended in a solution of 1 M NaCl or in PBS containing 0.1% 2-mercaptoethanol at 25°C x 30 min followed by centrifugation, separation of supernatant and pellet and comparison of the amount of immunoprecipitable Ly-S glyco- protein from each fraction. Lane A is a control supernatant from trypsinized cells. Lane B is an immunoprecipitate from non-trypsinized cells. Lane C is from pelleted trypsinized cells in which the Ly-S immunoprecipitate was solubilized in SDS sample buffer lacking Z-ME. Lane D is the same immunoprecipitate as C, except that the precipitate was solubilized in SDS sample buffer containing 2-ME. Lane E is the immunoprecipitable fraction of supernatant from pelleted trypsinized cells suspended in 1 M NaCl, while lane F is the corresponding cell- associated immunoprecipitate after NaCl treatment. Lane G represents solubilized Ly-S after treatment of pelleted cells with 0.1% 2-ME in PBS, with lane H indicating the immunoprecipitable Ly-S from the cell pellet after incubation with 2-ME. 101 116 — 92 — 67— I: Figure 6. Evidence that the 125 Kl) Ly-S fragment does not exist as a disulfide-linked heterodimer. Trypsinized 60% Percoll-fractionated cells were trypsinized and Ly-S immunoprecipitates solubilized in SDS sample buffer containing or lacking 2-mercaptoethanol. For designated lanes, preparative SDS-polyacrylamide gel electrophoresis was performed on an immunoprecipitate solubilized in SDS sample buffer without 2-ME. The gel was cut into 1 mm slices and the 125 KD fragment identified by counting slices in a gamma counter. Gel slices containing the 125 KD fragment were eluted into SDS sample buffer without 2-ME and divided into aliquots. To one sample was added 2-ME to a final concentration of 5%. Samples were heated to 100°C x 2 min and subjected to SDS-PAGE analysis. Lane A shows a standard Ly-S immunoprecipitate. Lane B is an Ly-S immunoprecipitate of trypsinized cells in SDS sample buffer without 2-ME. Lane C is an aliquot of the sample used in lane B to which 2-ME is added to a final concentration of 5% (v/v). Lane D is an immuno- precipitate from trypsinized cells directly solubilized in SDS sample buffer containing Z-ME. Lane B shows the electrophoretic mobility of the eluted 125 KD fragment in SDS sample buffer without 2-ME. Lane F is an aliquot of the same material in Lane E to which 2-ME is added to a final concentration of 5% (v/v). 102 presence of 2-ME (Figure 6, lane C), indicating that it did not exist as a heterodimer linked by a disulfide bond. This was further confirmed by elution of the 125 KD fragment from the nonreducing gel and repeat SDS- polyacrylamide gel electrOphoresis in the presence or absence of 2-ME (Figure 6, lanes E and F). Irypsinization of Ly-S from intact cells or' NP-HO solubilized membrane proteins yields fragments of identical apparent molecular mass. It has been established that Ly-S possesses an intracytoplasmic domain which can be labeled by BZPON and is extensively degraded by trypsin (20). It was of interest to determine whether the somewhat higher relative molecular mass and persistent membrane association of the Ly-S fragment observed in this study was due to the inclusion of the intra- membranous and cytoplasmic domains of the glycoprotein in fragments generated by trypsinization of intact cells. If true, the enzymatic digestion of these domains would be expected to occur only in NP-llo solubilized cells and result in a lower apparent Mr for the Ly-S frag- ment. However, SDS-PAGE analysis of Ly-S immunoprecipitates from intact cells and NP-MO solubilized membrane proteins subjected to trypsiniza- tion (Figure 7, lanes B and C, respectively), revealed that the Ly-S fragments generated had apparently identical Mr’ suggesting that trypsin generation of the fragment occurs entirely in the external domain. 103 ABCDEFGH 220 — 116 ”D. _ Figure 7. Apparent molecular weight identity of the 125 Kb Ly-5 fragment generated by trypsinization of intact cells and NP-flo solubilized cell surface proteins. Surface radioiodinated, 60% Percoll-fractionated splenocytes were trypsinized before and/or after solubilization in 1% NP-40 as described below. Ly—S immunoprecipitates and immunoprecipi- table SAC supernatants underwent SDS-polyacrylamide gel electrophoresis and autoradiography. Lane A is a standard Ly—S immunoprecipitate. Lane B represents the Ly-S fragment from cell surface proteins solubilized in NP-NO prior to trypsinization. Lane C is the Ly-S fragment generated by trypsin treatment of intact cells followed by the addition of Trasylol and PMSF before NP-MO solubilization. Lane D shows the Ly-5 fragment from intact cells treated with trypsin which were then solubilized in NP-HO with a delay in the addition of trypsin inhibitors for an additional 20 min. Lanes E, F, G and H are the nonimmunoprecipitable SAC supernatants from the starting material used in immunoprecipitation of samples used in Lanes A, B, C and D, respectively. 104 Discussion In a previous studyz, we employed specific antiserum and conjugate assays to show that Ly-5 glycoprotein is a mediator of murine natural killer cell binding to YAC-1 target cells. Using fluorescence-activated cell sorting, we further demonstrated that within murine splenic tissues are Ly-S pre-NK cells which can be induced by interferon and interleukin containing preparations to convert to Ly-S+ NK effectors. The acquired NK activity remained susceptible to inhibition by anti-Ly-S mAb, suggesting an important role for Ly-5/T200 glycoprotein in the natural killer cell cytotoxic mechanism against YAC-1 lymphoma targets. This is further supported by studies in the human natural killer cell system, where an antibody (13.1) to the human homologue of murine Ly-S/TZOO was shown to block the NK lytic mechanism between the stages of target binding and Ca++-dependent programming (22). Thus, separate determi- nants on a single glycoprotein have been implicated in successive stages of NK cell-mediated lysis, suggesting that Ly-S/TZOO glycoprotein acts as a NK cell receptor mediating the lytic reaction sequence. The experiments reported herein revealed several biochemical features of the Ly-S glycoprotein isolated from murine splenocyte suspensions enriched in NK cells and allowed comparison with available data from other studies regarding its disposition on the plasma membrane of other lymphocytes. A determination of the consistency of these bio- chemical properties with a current model of the mechanism of NK cell- mediated lysis (8) was also possible. 105 Based on SDS-PAGE analysis, an apparent approximate relative molecular mass of 205,000 was assigned to NK cell-associated Ly-5 glyco- protein. Its mobility when compared to whole splenocytes (Figure 2), and apparent Mr were indistinguishable from that of Ly-S glycoprotein isolated from macrophages (14). However, on some gels (Figure 3) the glycoprotein appeared to have a mobility similar to that of thymocytes. With its high degree of glycosylation (23), the broad bands observed on SDS-PAGE and the relatively small differences observed between macro- phages and T cells (111), an unequivocal association of murine NK cell surface Ly-S glycoprotein with a particular cell lineage is not possible. Such uncertainty is consistent with the results of previous studies in which NK cells were shown to have cell surface phenotypes and cellular functions resembling both macrophages and T cells (ZR-28). The observation that trypsin treatment of intact cells results in a immunOprecipitable fragment of Ly-S with approximate Mr of 125,000 that resists further trypsinization is similar to the results of Trowbridge et a1. (21), who identified a fragment of T200 on thymocytes with an apparent Mr of 100,000. Omary and Trowbridge (20), further characterized T200 from NP-llo solubilized membrane extracts of the murine lymphoma cell line BW51187 by surface radioiodination, metabolic labeling with 32m“, 3H--mannose or 3[SS-methionine and analysis of the effect of proteolysis on the glycoprotein. In surface radioiodinated cells, approximately 80% of the original T200 associated cpm was found in the 100 KD fragment generated by trypsin treatment. The fragment 106 also possessed most of the mannose-containing oligosaccharide portions of the glycoprotein. Analysis of isolated membrane preparations labeled by lactoperoxidase-catalyzed iodination revealed that the remainder of the molecule included a cytoplasmic domain which was extensively degraded by trypsin and contained phosphoserine residues that in viable cells could be labeled metabolically with 32P0”. Among the conclusions and implications of the data was that Ly-S/TZOO had an external domain containing most of the mannose-containing oligosaccharides and all of the amino acid residues labeled by radioiodination of viable cells, suggesting that the 100 KD fragment comprised most or all of the external domain of T200 and that the trypsin-sensitive peptide sequence was near the site of insertion of the glchprotein into the plasma membrane. In the absence of additional associations with the membrane or other membrane-bound molecules, the 100 KD fragment would be expected to become detached from the surface of trypsinized viable cells and appear in the surrounding medium. Our experimental observations that the 125 KD Ly-S fragment remained firmly bound to the plasma membrane after trypsinization of 'viable cells until solubilized by detergent suggested that the 125 KD fragment may contain the intramembranous and cytoplasmic portions of the glycoprotein. Additionally, the data suggested that the trypsin-sensitive site on the» external. domain. of Ly-S/T200 may reside closer to the distal end of the molecule than the membrane insertion site. This would be expected to create an additional peptide fragment of Ly-S with an expected Mr 0f approximately 75,000 bUt 107 which may not be detectable in these experiments due to further extensive degradation by trypsin, relative lack of amino acid residues susceptible or accessible to labeling by lactoperoxidase-catalyzed iodination or absence of antigenic determinants for Ly-S or T200 anti- sera necessary for immunoprecipitation and isolation of the peptide. This possibility was supported by a recent study in humans (29), in which a monoclonal antibody (9.1C3) that blocks NK cell activity was shown to precipitate 66 KD and 77 KD proteins associated with the T200 glycoprotein complex. Whether these peptides represent proteolytic degradation products of Ly-S/T200 has not been established. In some of the gels included in the current study (Figures 3, 6 and 7) a band was visualized at an apparent Mr of 85,000 KD but was considered to be a contaminating molecule, in that it was present in great abundance in nonimmunoprecipitates (M? both trypsinized and nontrypsinized samples, was not concentrated in Ly-S immunoprecipitates or substantially depleted from samples in which Ly-S had been removed by immunoprecipita- tion (data not shown). The possibility that the 85 KD band represented a portion of the external domain of the glycoprotein which lacks a Ly-S antigenic determinant could not be excluded. The 125,000 Mr of the trypsinized Ly-S fragments observed in the present experiments was somewhat greater than the 100,000 estimate reported. by (others (20,21). While procedural differences and broad banding pattern of the glycoprotein may have accounted for the observed differences, it was possible that trypsinization of Ly-S on intact cells 108 generated a larger peptide containing transmembrane and intracytoplasmic amino acid sequences not present in the 100,000 fragment generated by trypsinization of NP-llo solubilized Ly-S. The demonstrated suscepti- bility of the cytoplasmic portion of Ly-5/T200 to extensive degradation by trypsin (20) is in support of this postulate. However, direct molecular weight comparison of the immunoprecipitable Ly-S fragments generated by trypsin treatment of intact versus NP-llo solubilized cells and repeat trypsinization of the NP-HO solubilized fragment generated from intact cells failed to provide evidence for this, although the high degree of glycosylation and potentially small amount of intracytoplasmic peptide cleaved may have obscured evidence of differences. Repeat analysis after cytoplasmic domain labeling with 32P0u should more definitively answer this question and is currently under investigation. Recent studies have focused on Ly-S as a possible NK cell receptor participating in the mechanism leading to tumor cell lysis. Antisera reacting with different antigenic determinants associated with the 3137coprotein have been shown to block target cell bindingz, events linking target binding to Ca++-dependent programming (22) and post- ale"Jivation stages of NK cell-mediated lysis (29,30). Ly—S has also been Show to possess a transmembrane structure with a intracytoplasmic component (20), which has been postulated as a likely characteristic of an NK cell receptor (8). We have observed here that Ly-5 glycoprotein fr-Om NK-enriched splenocyte suspensions is highly susceptible to tr‘3'Dsinization, generating a fragment that resists further proteolytic 109 degradation and retains structural integrity sufficient to bind anti- Ly-s mAb. While our previous data2 suggest that the Ly-S determinant is important in NK cell binding to a lymphoma target, it appears unlikely that the Ly-S fragment is directly involved in the killer cell- independent lysis step of NK cell-mediated cytotoxicity in that it is not released from the plasma membrane by trypsinization. Although the ability of membrane-bound Ly-5 fragment to mediate continued binding to target cells has not been established, it is likely that any in v_iv_g target cell binding mediated by the membrane-bound fragment would be disrupted by endogenous proteases responsible for its generation. Such proteolysis would be expected to lead to cleavage of receptors on the target cell, which has been shown to block their susceptibility to binding and lysis by NK effectors (5,7,8). The possibility that proteo- lysis of Ly-5 glycoprotein or a closely associated molecule liberates a peptide from the cell surface capable of participating in the lethal hit stage of NK cell-mediated lysis is suggested by recent reports (29,30), but requires further investigation. In attributing unique biological properties to NK cell-associated Ly~5 glycoprotein, one must address the fact that the expression of Ly-S is not unique to these effector cells. The molecular weight hetero- g‘el'leity observed for Ly-S/TZOO (13,111,31) suggests structural variations an“Orig cells of different hematopoietic lineage that may alter the 1‘=."4‘1'14:‘.~tion of the glycoprotein. An example of documented differences in the structure of the glycoprotein was provided by Coffman and Weisman 110 (32), who developed a monoclonal antibody that reacted specifically with the 220,000 dalton form of T200 that is present on B cells, which they designated 8220. Other studies have provided additional evidence of differences between T and B cells in the protein structure of Ly-S and its intracytoplasmic precursors (33,3H). Variation in the oligosaccha- ride components of Ly-S among different lymphocyte populations as a contributing factor for Ly-S heterogeneity awaits forther study. Acknowledgements: We wish to thank M. Morrison and J. Wildenthal .for their valuable technical assistance in these experiments. References 1 . Herberman, R. B. (1981) Natural killer (NK) cells and their possible roles in resistance against disease. Clin. Immunol. 333. 1, 1. 22‘. Skamene, E., Stevenson, M. M., and Kongshavn, A. L. (1982) Natural cell-mediated immunity against bacteria, p. 1513. In NK Cells and Other Natural Effector Cells, edited by R. Herberman, Academic Press, Inc., New York, NY. 33 - Kiessling, R., Hochman, P. S., Haller, 0., Shearer, G. M., Wigzell, H., and. Cudkowicz, G. (1977) Evidence for* a similar or common mechanism for natural killer cell activity and resistance to hemopoietic grafts. Eur. J. Immunol. 7, 655. 111 u. Hansson, M., Kiessling, R., and Beran, M. (1982) Natural killing of“ hematopoietic cells, p. 1077. In NK Cells and Other Natural Effector Cells, edited by R. Herberman, Academic Press, Inc., New York, NY. 5. Roder, J. C., Kiessling, R., Biberfeld, P., and Anderson, B. (1978) Target-effector interaction in the natural killer cell system. II. The isolation of’ NK cells and studies on the mechanism of killing. J. Immunol. 121, 2509. 6. Hiserodt, J. C., Britvan, L. J., and Targan, S. R. (1982) Charac- terization of the cytolytic reaction mechanism of the human natural killer (NK) lymphocyte: resolution into binding, programming and killer cell-independent steps. .L. Immunol. 129, 1782. 7. Hiserodt, J. C., Britvan, L. J., and Targan, S. R. (1983) Studies on the mechanism of the human natural killer cell lethal hit: evidence for the transfer of protease-sensitive structures requisite for target cell lysis. J. Immunol. 131, 2710. 8. Hiserodt, J. C., Britvan, L. J., and Targan, S. R. (1983) Studies on the mechanism of the human natural killer cell lethal hit: analysis of the mechanism of protease inhibition of the lethal hit. J. Immunol. 131, 2705. 9. Minato, N., Reid, L., Cantor, H., Lengyel, P., and Bloom, B. R. (1980) Mode of regulation of natural killer cell activity by interferon. J. Exp. Med. 152, 124. 10. 11. 12. 13. 111. 15. 16. 112 Kasai, M., Leclerc, J. C., Shen, F. W., and Cantor, H. (1979) Identification of Ly-S on the surface of 'natural killer' cells in normal and athymic inbred mouse strains. Immuno- genetics 8, 153. Pollack, S. B., Tan, M. R., Nowinski, R. C., and Emmons, S. L. (1979) Presence cu“ T-cell associated surface antigens on murine NK cells. J. Immunol. 123, 1818. Siadek, A. W., and Nowinski, R. C. (1980) Identification of Ly-S and T200 antigens on identical cell surface proteins. _J_._ Immunol. 125, 1HOO. Trowbridge, I. A., Ralph, P., and Bevan, M. J. (1975) Differences in the surface proteins of mouse B and T cells. Proc. Natl. Acad. Sci. USA 72, 157. Tung, J.-S., Scheid, M. P., Pierotti, M. A., Hammerling, U., and Boyse, E. A. (1981) Structural features and selective expression of three Ly-5+ cell surface molecules. Immuno- genetics 1", 101. Esselman, W. J., and Miller, H. C. (1977) Modulation of B cell response by glycolipid released from antigen stimulated T cells. J. Immunol. 119, 199". Nabel, G., Fresno, M., Chessman, A., and Cantor, H. (1981) Use of cloned pOpulations of mouse lymphocytes to analyze cellular differentiation. Cell 23, 19. 113 17. Kurnick, J. T.,‘Ostberg, L., Stegagno, M., Kimura, A. K., Orn, A., and Sjoberg, O. (1979) 11 rapid method for the separation of functional lymphoid cell populations of human and animal origin on PVP-silica (Percoll) density gradients. Scand. J. Immunol. 10, 563. 18. Morrison, M. H., Chaney, W. G., and Esselman, W. J. (198”) Molecular weight and charge heterogeneity of Thy-1 glyco- proteins in different populations of T-cells. Molec. Immunol. 21, 905. 19. Tagliabue, A., Boraschi, D., Alberti, S., and Luini, W; (1982) Large granular lymphocytes as effector cells of natural killer cell activity in the mouse, p. 25. In NK Cells and Other Natural Effector Cells, edited by R. Herberman, Academic Press, Inc., New York, NY. 20. Omary, M. B., and Trowbridge, I. S. (1980) Disposition of T200 glycoprotein in the plasma membrane lymphoma cell line. .J; Biol. Chem. 255, 1662. 21. Trowbridge, I. S., Nelsen-Hamilton, M., Hamilton, R. T., and Bevan, M. .J. (1977) Preliminary' characterization of’ two 'thymus- dependent xenoantigens from mouse lymphocytes. Biochem. J. 163, 211. 22. Targan, S. R., and Newman, W. (1983) Definition of a "trigger" stage in the NK cytolytic reaction sequence by a monoclonal antibody to the glycoprotein T-200. J. Immunol. 131, 11u9. 23. 2“. 25. 26. 27. 28. 29. 114 Watson, A., Dunlap, B., and Bach, F. H. (1981) The biosynthesis of Ly-5 in T and B cells. J. Immunol. 127, 38. Herberman, R. B., and Holden, H. T. (1978) Natural cell-mediated immunity. Adv. Cancer Res. 27, 305. Meruelo, D., Paolino, A., Flieger, N., and Offer, M. (1980) Definition of a new T lymphocyte cell surface antigen, Ly 11.2. J. Immunol. 125, 2713. Archer, J. R., and Davies, A. L. (1974) Demonstration of an allo- antigen on the peritoneal exudate cells of inbred strains of mice and its association with chromosome 7 (linkage group I). J. Immunogen. 1, 113. Springer, T., Galfre, G., Secher, D. S., and Milstein, C. (1978) Monoclonal xenogeneic antibodies to murine cell surface anti- gens: identification of novel leukocyte differentiation anti- gens. Eur. J. Immunol. 8, 539. Kasai, M., Iwamori, M., Nagai, Y., Okumura, K., and Tada, T. (1980) A glycolipid in the surface of mouse natural killer cells. Eur. J. Immunol. 10, 175. Burns, G. F., Wenkmeister, J. A., and Triglia, T. (1984) A novel antigenic cell surface protein associated with T200 is involved in the post-activation stage of human NK cell- mediated lysis. J. Immunol. 133, 1391. 30. 31. 32. 33. 34. 115 werkmeister, J. A., Burns, G. F., and Triglia, T. (1984) Anti- idiotype antibodies to the 9.103 blocking antibody used to probe the lethal-hit stage of NK cell-mediated cytolysis. J; Immunol. 133, 1385. Michaelson, J., Scheid, M., and Boyse, E. A. (1979) Biochemical features of Ly-S alloantigen. Immunogenetics 3, 193. Coffman, R. L., and Weisman, I. L. (1981) B220: a B cell- specific member of the T200 glycoprotein family. Nature 289, 681. Tung, J.-S., Meere, M. C., and Boyse, E. A. (1984) Evidence that Ly-5‘product of T and B cells differ in protein structures. Immunogenetics 19, 149. Watson, A., Dunlap, B., and Bach, F. H. (1981) The biosynthesis of Ly-5 in T and B cells. J. Immunol. 127, 38. SUMMARY AND CONCLUSIONS In these studies, inquiry was made into fundamental aspects of murine natural killer cell development and function. Using specific antiserum to inhibit NK cell function, Ly—S/TZOO glycoprotein was identified as a candidate cell surface molecule important in NK cell-mediated cytotoxicity. Monoclonal anti-Ly-S was used to probe the stage in the lytic mechanism in which the Ly-S determinant participates. The results of target cell binding assays indicated that the Ly-S determinant or a closely related structure is involved in the Mg++-dependent binding stage of the NK cell lytic mechanism against YAC-l lymphoma target cells. The importance of Ly-S to NK cell function was further supported by the finding that cell sorter purified splenocytes lacking significant amounts of surface Ly-5 glycoprotein were without NK activity. One possible mechanism by which the endogenous immunomodulators interferon and interleukin-2 augment natural killer cell activity was identified, in that medium containing these agents was observed to increase the surface Ly-S expression and NK cell activity of sorted Ly-5- cells. Prolonged in vitro culture in medium containing IFN and IL—2 was 116 117 associated with at least a tenfold increase in surface Ly-S expression and potent NK cell activity that remained significantly inhibitable by anti-Ly-S mAb. Once evidence for the importance of Ly-5 in murine NK cell development, target binding and cytotoxicity was established, it was of interest to gain insight into the biochemistry of the glycoprotein on NK cells to allow comparison with other hematopoietic cells. The relative molecular mass of the glycoprotein obtained from NK cell-enriched spleen cell suspensions was 205,000 and was indistinguishable from the Mr obtained for macrophages. The results from proteolysis experiments indicated that trypsinization of intact cells did not alter the amount of immunofluorescently detectable Ly-5 on the plasma Inembrane. Immunoprecipitation and SDS—polyacryalamide gel electrOphoretic analysis of surface radioiodinated cells :revealed that the glycoprotein was cleaved to a 125 KD fragment that remained bound to the plasma membrane and (was not liberated into the soluble phase by treatment with sinmde sugars, high salt or disulfide inhibitors. The iiragment was resistant to further degradation by trypsin arui retained its ability to bind anti-Ly-S mAb. At least sonma of these properties are consistant with current modeflls of the properties of a putative NK cell receptor. Taken together, the findings described in this dissexrtation, along with the results of others in the humarl and murine natual killer cell systems, suggest an 118 important and fundamental role for Ly-S/TZOO in the NK cell-mediated cytotoxic mechanism against tumor cells. M1111111111[111111111111111” 1 9 6796