Antiviral effect of lymphokine-activated killer cells: chemotaxis and homing to sites of virus infection.

Lymphokine-activated killer (LAK) cells generated from C57BL/6 mouse spleen cells cultured with interleukin-2 are effective prophylactically against virus infection when inoculated at the site of virus injection. To predict the therapeutic efficacy of LAK cells, we determined whether LAK cells would home to sites of virus infection. In vitro, LAK cells responded chemotactically to cell-free peritoneal exudate fluids collected from virus-infected mice and to preparations of purified beta interferon. In vivo, radiolabeled LAK cells injected intravenously accumulated in the peritoneal cavities of intraperitoneally infected mice in amounts three to eight times greater than in uninfected mice. This ability to respond to chemotactic agents and migrate into sites of virus infection may make LAK cells useful as antiviral therapeutic agents.     

Lysis of mycobacteria-infected monocytes by IL-2-activated killer cells: role of LFA-1

Mycobacterium avium-intracellulare (MAI) is a ubiquitous soil contaminant that rarely causes disseminated disease in adults regardless of immunological status. In AIDS patients, however, this organism invades virtually every tissue and organ, and most conventional chemotherapeutic agents are usually ineffective against MAI. We report here that monocytes, in which MAI has established an intracellular parasitic stage, are under the control of natural killer (NK) cells. Autologous large granular lymphocytes (LGL), purified from human peripheral blood leukocytes, were capable of efficiently lysing autologous MAI-infected monocytes in a 5-hr 51Cr release assay. More importantly, interleukin 2 (IL-2) was able to activate the LGL to a higher degree of lysis of infected monocytes. LGL cultured in medium alone could not kill normal monocytes, but showed some degree of lysis of MAI-infected cells. IL-2 activated killer (LAK) cells, on the other hand, lysed normal monocytes to a moderate degree and this activity was makedly enhanced if the monocytes were infected with MAI. The sensitivity of monocytes was directly proportional to the inoculating number of bacteria, indicating that increased bacterial burden would enhance susceptibility to LAK-mediated lysis. Finally, the addition of monoclonal antibodies to LFA-1 (both alpha and beta chains), but not LFA-2 or LFA-3, blocked lysis of both infected and uninfected monocytes when added directly to the cytotoxicity assays, indicating that this adhesion protein is involved in the lysis of autologous, infected monocytes. Thus, NK/LAK cells may be important in containment of infection by lysis of infected monocytes before the bacteria can multiply and spread to other sites.     

Generation of Cyclic ADP-ribose and Nicotinic Acid Adenine Dinucleotide Phosphate by CD38 for Ca2+ Signaling in Interleukin-8-treated Lymphokine-activated Killer Cells

We have previously demonstrated that cyclic ADP-ribose (cADPR) is a calcium signaling messenger in interleukin 8 (IL-8)-induced lymphokine-activated killer (LAK) cells. In this study we examined the possibility that IL-8 activates CD38 to produce another messenger, nicotinic acid adenine dinucleotide phosphate (NAADP), in LAK cells, and we showed that IL-8 induced NAADP formation after cADPR production. These calcium signaling messengers were not produced when LAK cells prepared from CD38 knock-out mice were treated with IL-8, indicating that the synthesis of both NAADP and cADPR is catalyzed by CD38 in LAK cells. Application of cADPR to LAK cells induced NAADP production, whereas NAADP failed to increase intracellular cADPR levels, confirming that the production of cADPR precedes that of NAADP in IL-8-treated LAK cells. Moreover, NAADP increased intracellular Ca²⁺ signaling as well as cell migration, which was completely blocked by bafilomycin A1, suggesting that NAADP is generated in lysosome-related organelles after cADPR production. IL-8 or exogenous cADPR, but not NAADP, increased intracellular cAMP levels. cGMP analog, 8-(4-chlorophenylthio)-guanosine 3′,5′-cyclic monophosphate, increased both cADPR and NAADP production, whereas the cAMP analog, 8-(4-chlorophenylthio)-cAMP, increased only NAADP production, suggesting that cAMP is essential for IL-8-induced NAADP formation. Furthermore, activation of Rap1, a downstream molecule of Epac, was required for IL-8-induced NAADP formation in LAK cells. Taken together, our data suggest that IL-8-induced NAADP production is mediated by CD38 activation through the actions of cAMP/Epac/protein kinase A/Rap1 in LAK cells and that NAADP plays a key role in Ca²⁺ signaling of IL-8-induced LAK cell migration.     

Modulation of perforin, granzyme A, and granzyme B in murine natural killer (NK), IL2 stimulated NK, and lymphokine-activated killer cells by alcohol consumption.

Alcohol consumption in mice suppresses the cytolytic activity of natural killer (NK) and lymphokine-activated killer (LAK) cells through unknown mechanisms. Herein, we found that alcohol consumption decreased target cell-induced release of granzyme A activity in freshly isolated splenic NK cells, in NK cells stimulated for 18 h with 1000 IU/ml of interleukin 2, and in LAK cells. The total activity and protein expression of granzymes A and B also were lower in these cells than in cells isolated from water-drinking mice. Interleukin 2 increased granzyme A protein expression independent of alcohol consumption; however, this increase was associated with decreased enzyme activity. In contrast, granzyme B protein expression and enzymatic activity increased in response to interleukin 2. Perforin activity and protein expression were reduced in LAK cells generated from alcohol-consuming mice. We conclude that the mechanism underlying the suppression of NK and LAK cytolytic activity by alcohol consumption involves the collective reduction of target-induced release, activity, and expression of perforin and granular proteases.     

Assessment of human natural killer and lymphokine-activated killer cell cytotoxicity against Toxoplasma gondii trophozoites and brain cysts

Because previous work has suggested that NK cells may be important in host resistance against the intracellular parasite Toxoplasma gondii we examined whether human NK cells and lymphokine-activated killer (LAK) cells have activity against trophozoites and cysts of this organism in vitro. A method to radiolabel Toxoplasma trophozoites with 51Cr was developed and direct cytotoxic activity was determined by using modifications of the standard 51Cr release assay. Viability of 51Cr-labeled trophozoites assessed by both methylene blue staining and trypan blue exclusion was greater than 90%. Significantly more 51Cr was released by anti-Toxoplasma antibody and C than by antibody in the absence of C. Incubation of trophozoites with freshly isolated human NK cells or NK cells activated with either rIL-2 or rIFN-alpha did not result in significant release of 51Cr (specific lysis was 0 to 2.3%). In contrast, the average specific lysis of radiolabeled trophozoites by LAK cells was significant (specific lysis was 7.8% +/- 1.1, p less than 0.01). In a series of separate experiments, preincubation of radiolabeled trophozoites with heat-inactivated normal or Toxoplasma antibody-positive human serum increased the cytotoxicity of LAK cells from a mean specific lysis of 15% +/- 4.5 to 39% +/- 8.5, respectively (p less than 0.05), as assessed by 51Cr release. Because previous work has shown that radioisotope release from parasites may be nonspecific, separate experiments were performed to determine the cytotoxicity of LAK cells against antibody-coated trophozoites by using ethidium bromide-acridine orange staining to assess effector cell damage. LAK cells had a mean specific lysis of 51% against antibody-coated trophozoites by ethidium bromide-acridine orange staining. Preincubation with heat-inactivated Toxoplasma-antibody positive human serum did not increase activity of rIL-2-activated NK cells against 51CR-labeled trophozoites. Neither human NK cells (freshly isolated or activated by rIL-2 or rIFN-alpha) nor LAK cells were cytotoxic for purified preparations of cysts of Toxoplasma isolated from the brains of chronically infected mice.     

Soluble factors produced by macrophages/monocytes inhibit lymphokine-activated killer activity in rat splenocyte cultures

In the present study we investigated the inhibition of interleukin-2(IL-2)-induced lymphokine-activated killer (LAK) activity in rat splenocyte cultures in relation to the presence of 2-mercaptoethanol and macrophages/monocytes. The presence of 2-mercaptoethanol is necessary for induction of LAK activity in rat splenocyte cultures. Removal of macrophages/monocytes from rat splenocytes by plastic or nylon-wool adherence, or iron ingestion resulted in LAK induction by IL-2 in the absence of 2-mercaptoethanol. The effect of macrophages/monocytes on LAK activity was also studied in transwell co-cultures. In the absence of 2-mercaptoethanol, the induction of LAK activity was very low in macrophage/monocyte-depleted splenocytes with macrophages/monocytes in the upper compartment of a transwell culture. In contrast, in the presence of 2-mercaptoethanol a high level of LAK activity was induced in these transwell cultures, showing that 2-mercaptoethanol abolished the LAK-inhibiting capacity of macrophages/monocytes. In addition, established LAK activity was strongly inhibited when, after LAK induction, splenocytes were cultured with supernatant of unfractionated splenocytes, which were cultured with IL-2 but in the absence of 2-mercaptoethanol. Addition of 2-mercaptoethanol abrogated the inhibiting effect of the supernatant completely. These experiments demonstrate that rat macrophages/monocytes produce 2-mercaptoethanolsensitive soluble LAK-inhibiting factors. Ultrafiltration of conditioned culture medium of macrophages/monocytes revealed the presence of LAK-inhibiting factors larger than 10 kDa. We concluded that 2-mercaptoethanol-sensitive soluble factors produced by macrophages/monocytes determine the level of LAK induction in rat splenocyte cultures.     

Effects of human alveolar macrophages on the induction of lymphokine (IL 2)-activated killer cells

Normal human alveolar macrophages (AM) significantly and reproducibly suppress induction of IL 2-activated killer (LAK) cell activity against allogeneic Burkitt's lymphoma (Daudi) cells. Incubation of purified peripheral blood lymphocytes for 4 days with autologous AM and 1 U/ml of IL 2 resulted in AM-mediated suppression of LAK activity, whereas peripheral blood monocytes isolated freshly by centrifugal elutriation from the same donor potentiated induction of LAK activity by IL 2. The suppression of LAK cell induction by human AM was dependent on the density of AM added to the lymphocyte cultures. Recombinant IFN-gamma did not affect AM-mediated suppression of LAK cell induction by IL 2. Both AM and monocytes stimulated with lipopolysaccharide markedly suppressed LAK cell induction by IL 2. AM-mediated down-regulation was seen only when AM were added immediately after the start of incubation of lymphocytes with IL 2; AM potentiated LAK activity when added 1 day later. Similar AM-mediated suppression of LAK cell induction was observed with four lines of allogeneic lung cancer cells as targets for LAK activity. These results indicate that AM may be important in regulation of in situ induction of LAK activity in the lung.     

Immunosuppression in murine renal cell carcinoma

In our companion paper we have reported that cell-mediated immunity of mice bearing renal cell carcinoma is profoundly suppressed. The non-responsiveness of such animals was found to be attributable to Renca cells themselves and to splenic lymphoid cells that down-regulate other fully capable lymphoid cells. In this communication the lymphoid cell source of suppression within Renca-bearing mice has been explored with the aim of identifying phenotypes of the responsible cells, the manner by which suppression is mediated, and initial ways by which suppression may be eliminated. A plastic-adherent cell bearing the Thy1.2 surface marker as well as the Lyt1 and Lyt2 antigens has been found to operate, perhaps in conjunction with macrophages, to down-regulate lymphokine-activated killer (LAK) cell development for natural killer (NK) and non-NK targets that include Renca cells themselves. The splenic suppressor cells lost the capacity to suppress the NK response of normal recipient mice upon shallow irradiation (250 rad) prior to adoptive transfer. Spleen cells, presumably macrophages, from Renca-bearing mice were found to suppress the generation of LAK and NK cells in vitro by synthesizing prostaglandins. Indomethacin, a prostaglandin synthetase inhibitor, blocked the induction of suppression both in vitro and in vivo, suggesting the presence of endogenous prostaglandins in Renca-bearing mice. The suppression seen in Renca-bearing mice that derives from multiple sources and has been prevented by two separate methods has been discussed from the viewpoint of the inter-relatedness of the sources.     

Lymphokine-activated killer cells in rats. IV. Developmental relationships among large agranular lymphocytes, large granular lymphocytes, and lymphokine-activated killer cells

The developmental relationships among large agranular lymphocytes (LAL) large granular lymphocytes (LGL) and the activation of these cells into lymphokine-activated killer (LAK) cells by rIL-2 was investigated. Highly enriched populations of LAL were isolated from Fischer 344 spleen cells by a combination of nylon-wool filtration (to remove B cells and macrophages), treatment with a pan T cell antibody plus complement (to remove T cells) and incubation in L-leucine methyl ester (to remove LGL). The resultant cells were highly enriched in morphologically identifiable LAL which expressed asialo GM1 and partially expressed the OX8 surface marker. The enriched LAL did not contain detectable NK cytotoxic activity, did not express pan T cell (OX19), Ia, Ig, or laminin surface markers and contained less than 0.2% LGL. Incubation of LAL in a low dose of rIL-2 (100 U/ml) induced the generation of LGL having NK activity within 24 h of culture. Longer culture periods (48 h) resulted in a continued increase in the percentage of LGL and higher levels of NK activity. However, with this low dose of rIL-2, little or no LAK activity (i.e., reactivity against NK-resistant target cells) was generated. With a high dose of rIL-2 (500 U/ml), LAL responded by first generating LGL with NK activity (within 24 h), with subsequent generation of LAK activity by 48 h. Evidence that the development of granular lymphocytes from LAL was responsible first for NK activity and then LAK activity was demonstrated by depletion of the generated granular NK or LAK effector cells by second treatments with L-leucine methyl ester. Concomitant with the induction of LGL with NK or LAK activity, rIL-2 also caused LGL to proliferate and expand four- to five-fold in 48 h. This occurred in the presence of high or low dose rIL-2. These results indicate that LAL are the precursors of LGL/NK cells, that LAL, LGL/NK cells and LAK cells appear to represent sequential developmental or activation stages and that LAL may comprise major source of LAK progenitors in lymphoid populations having few LGL or mature active NK cells.     

Effector and precursor phenotypes of lymphokine-activated killer cells in mice with severe combined immunodeficiency (scid) and athymic (nude) mice

The lineage of lymphokine-activated killer (LAK) cells is poorly understood. To examine the relationship between LAK and natural killer (NK) cells we utilized two congenitally immunodeficient mice, namely severe combined immunodeficient (scid) and athymic (nude) mice that lack T cells but have normal NK cells. LAK activity was evaluated by the ability to lyze NK-resistant P815 cells. When cultured with human recombinant interleukin 2, splenocytes of scid and nude mice could generate LAK activity at levels comparable to or more than those of normal C.B-17 mice. LAK effector cells in these immunodeficient mice as well as normal mice had the phenotype resembling that of NK cells with asialo-GM1 (aGM1) expression. In vivo treatment with anti-aGM1 antiserum completely abolished the induction of LAK activity from splenocytes of normal mice. In contrast, LAK activity in splenocytes of scid and nude mice was still demonstrable even after this treatment, indicating that most LAK precursors in both mice were cells without aGM1 antigen. The aGM1- progenitors for LAK activity, probably in common with NK progenitors, appeared to be more expanded in scid and nude mice than in normal mice. The use of such congenitally immunodeficient mice should be helpful in studying the differentiation step of LAK as well as NK cells from their precursors.     

Lymphokine-activated killer cells are rejected in vivo by activated natural killer cells

A 4-h in vivo cytotoxicity assay was used to study the fate of implanted IL-2-generated, lymphokine-activated killer (LAK) cells in mice undergoing an activated NK cell response. 125Iododeoxyuridine-labeled LAK cells were rejected from selected organs of C57BL/6 mice infected with lymphocytic choriomeningitis virus or treated with IL-2 or the IFN inducer poly I:C. This rejection was abrogated by the selective depletion of NK cells with antibodies to asialo-GM1 and NK1.1 Ag. Similar results were noted when LAK cells were generated from the spleens of B and T cell-deficient severe combined immunodeficiency mice and when LAK cells were implanted into severe combined immunodeficiency mice. These data indicate that NK cells activated by virus infections or by IL-2 infusions directly or indirectly eliminate implanted LAK cells. Because LAK cells are used in the treatment of certain human cancers, the strategy of accompanying this therapy with IL-2 infusions should be reassessed in light of these results.     

Lymphokine-activated cell-associated antigen involved in broad-reactive killer cell-mediated cytotoxicity

Spleen cells from rats which had been hyperimmunized with mouse lymphokine-activated killer (LAK) cells, were fused with the mouse myeloma cell line, P3 X 63 Ag8.653. Antibodies secreted by 1500 cultures were selected by their blocking effect on LAK cell-mediated cytotoxicity in the absence of complement. Two monoclonal antibodies (KBA4 and KBA6) greatly inhibited the cytotoxic activity of LAK cells, which were induced from mouse spleen cells by culture with recombinant human interleukin 2 (r-IL-2). These antibodies also blocked the cytotoxic activity of natural killer (NK) cells, but activated macrophages (A-M phi) were only slightly sensitive to them. However, no effect of the antibodies on the cytotoxic activity of cytotoxic T lymphocytes (CTL) was detected. These data suggest that the specific antigen, lymphokine-activated cell-associated (LAA) antigen, defined by these monoclonal antibodies may be associated with the recognition mechanisms of broad-reactive killer (BRK) cell-mediated cytotoxicity. The observation that low levels of LAA antigen are distributed in all lymphoid cells and that it was significantly enhanced by treatment of the cells with r-IL-2 suggests that the antigen may be involved in lymphocyte-activation mechanisms. We also found that the LAA antigen consists of two distinct polypeptides with Mr of 180,000 and 95,000 Da, which are similar to that of LFA 1 antigen. However, the biological characteristics of LAA antigen did not coincide with those of LFA 1. Therefore, KBA MAb may recognize a carbohydrate epitope distinct from that of LFA 1.     

In vivo generation of lymphokine-activated killer cells by sensitization with interleukin 2-producing syngeneic T-lymphoma cells

Culture of C57BL/6 mouse spleen cells with syngeneic EL4 lymphoma cells resulted in no induction of killer cells reactive against EL4 cells. However, in vitro sensitization of C57BL/6 mouse spleen cells with interleukin 2 (IL-2)-producing EL4 lymphoma cells caused the generation of lymphokine-activated killer (LAK) cells, which lyse a variety of tumor cells. Consistent with an in vitro system, we demonstrate that Thy 1.2+, Ly2+, asialo GM1+ LAK cells were successfully induced by in vivo immunization with syngeneic IL-2-producing EL4 lymphoma cells.     

Association of CD38 with nonmuscle myosin heavy chain IIA and Lck is essential for the internalization and activation of CD38

Activation of CD38 in lymphokine-activated killer (LAK) cells involves interleukin-8 (IL8)-mediated protein kinase G (PKG) activation and results in an increase in the sustained intracellular Ca(2+) concentration ([Ca(2+)](i)), cADP-ribose, and LAK cell migration. However, direct phosphorylation or activation of CD38 by PKG has not been observed in vitro. In this study, we examined the molecular mechanism of PKG-mediated activation of CD38. Nonmuscle myosin heavy chain IIA (MHCIIA) was identified as a CD38-associated protein upon IL8 stimulation. The IL8-induced association of MHCIIA with CD38 was dependent on PKG-mediated phosphorylation of MHCIIA. Supporting these observations, IL8- or cell-permeable cGMP analog-induced formation of cADP-ribose, increase in [Ca(2+)](i), and migration of LAK cells were inhibited by treatment with the MHCIIA inhibitor blebbistatin. Binding studies using purified proteins revealed that the association of MHCIIA with CD38 occurred through Lck, a tyrosine kinase. Moreover, these three molecules co-immunoprecipitated upon IL8 stimulation of LAK cells. IL8 treatment of LAK cells resulted in internalization of CD38, which co-localized with MHCIIA and Lck, and blebbistatin blocked internalization of CD38. These findings demonstrate that the association of phospho-MHCIIA with Lck and CD38 is a critical step in the internalization and activation of CD38.     

Adoptive immunotherapy of malignant diseases with IL-2-activated lymphocytes

Lymphokine activated killer cells (LAK cells) or interleukin 2 (IL-2)-activated killer cells were induced by recombinant IL-2 (TGP-3) for clinical adoptive immunotherapy of malignant diseases. After incubation of peripheral blood lymphocytes (PBL) with IL-2 and normal human plasma for 1-2 weeks LAK cells were obtained that showed a maximum cytotoxicity against target cells, and did not need a toxic dose of IL-2 to enhance or maintain their cytotoxicity. Both autologous and allogeneic LAK cells were used in five clinical cases without any immune side effects, and were effective in three cases.     

Characterization of effector-target conjugates for cloned human natural killer and human lymphokine activated killer cells by flow cytometry.

The present studies demonstrate that the intracellular fluorochromes calcein and hydroethidine can be used for quantification of effector-target conjugates involving cloned human natural killer (NK) or interleukin-2 (IL-2) activated human lymphokine activated killer (LAK) cells by dual color flow cytometry without potential artifacts that might result from extensive modification of effector and/or target cell membranes. Cloned NK cells and LAK cells form conjugates with cultured cell lines regardless of susceptibility to lysis. The strength of the interactions in these conjugates was investigated using a variable speed vortexer. Even relatively gentle vortexing disrupted most conjugates involving fresh human peripheral blood lymphocytes (PBL) but only about one-fourth of conjugates between K-562 cells and human PBL that had been cultured with or without IL-2 by this treatment. The rate of conjugate formation for LAK cells was determined to be about 3 times faster than for cloned NK cells, and both rates are considerably faster than the reported rate of formation of cytotoxic T lymphocyte (CTL) target conjugates. The differences in the rate of conjugate formation are apparently not related to target cell specificity, since LAK cells form conjugates with susceptible and resistant cell lines at comparable rates. When effector-target conjugates are incubated at 37 degrees C in the absence of calcium--thereby precluding lysis--the percentage of conjugated LAK or cloned NK cells decreases logarithmically with time. These results suggest that an initial equilibrium between free and conjugated lymphocytes gradually shifts in favor of unconjugated cells.     

Modification of lymphokine-activated killer cell accumulation into tumor sites by chemotherapy, local irradiation, or splenectomy

The effects of chemotherapy or local irradiation on lymphokine-activated killer (LAK) cell accumulation into tumor sites were investigated. Lymphokine-activated killer cells labeled with 111In-oxine were injected into the caudal vein of C57BL/6 mice that had been previously transplanted with 3LL cancer. An adoptive transfer of LAK cells was carried out 4 days after treatments. Twenty-four hours after the transfer, tumor tissues were excised, and the accumulation of labeled LAK cells in the tumor was measured. In two different experiments, LAK cell accumulation in tumor in the nontreated group was 2.15% and 1.58% of the administered dose per gram of tissue. The accumulation in the groups of mice treated with cyclophosphamide, nimustine hydrochloride, or Adriamycin increased fourfold (7.38% dose/g, 6.61% dose/g), threefold (6.47% dose/g) and twofold (4.46% dose/g), respectively, as compared with the nontreated group. These agents induced significant tumor regression. In the group treated with bleomycin, which showed no significant effect on tumor growth, LAK cell accumulation in tumor remained unaltered (1.57% dose/g). However, the group treated with local irradiation, which induced significant tumor reduction, showed no increase in LAK cell accumulation into tumors. These results suggest that some antitumor drugs enhance LAK cell accumulation into tumor sites and that this increase is due to tumor modification by antitumor drugs.     

Lymphokine-activated killer (LAK) cells. II. Delineation of distinct murine LAK-precursor subpopulations

Lymphokine-activated killer (LAK) cells can lyse a number of tumor target cells regardless of whether the tumors are natural killer (NK) sensitive or resistant. LAK can also lyse autologous lymphoblasts that have been modified with 2,4,6-trinitrobenzene sulfonic acid (TNBS). In this study, we examined the surface markers of murine LAK precursors. It was found that depletion of Thy 1- or Lyt 2-bearing precursor cells abolished the ability of spleen cells to generate LAK against TNBS-self, but had no effect on the generation of LAK against tumor cells. Depletion of asialo-GM1 (AGM1)-bearing precursors abolished the generation of LAK against all target cells tested. Normal spleen cells were fractionated on a Percoll density gradient and two fractions were examined: fraction (Fxn) 3, which is enriched for NK activity but depleted of the ability to generate cytotoxic T lymphocytes (CTL), and Fxn 5, which had no NK activity but was enriched for the ability to generate CTL. Both fractions were capable of generating LAK, although Fxn 5 required a relatively larger amount of interleukin 2 (IL 2). Upon examination of the surface markers of LAK precursors in these fractions it was found that the precursors in Fxn 3 giving rise to LAK against tumors were Thy-1-, Lyt-2-, AGM1+, whereas the precursors in Fxn 5 were Thy-1+, Lyt-2+, AGM1+. The precursors generating LAK against TNBS-self were Thy-1+, Lyt-2+, AGM1+ in both fractions. The time kinetics of LAK generation in both fractions were different, with Fxn 3 showing much earlier kinetics. These data delineate at least two different LAK precursors defined by their buoyant density, by their surface markers, and by their susceptible target cells. These data also may resolve the confusion in the literature regarding the phenotype of LAK precursors.