NK cells are most effective in killing a broad spectrum of primary tumor cells after stimulation with cytokines. We have cloned a novel gene, designated NKLAM (for NK lytic-associated molecule), whose expression is associated with this cytokine-enhanced process. NKLAM expression is up-regulated in NK cells by IL-2 and IFN-β. NKLAM is also selectively expressed by activated macrophages and CTL. Treatment of NK cells and CTL with NKLAM antisense oligonucleotides specifically decreases their cytolytic activity, while having no effect on cell growth. The NKLAM gene encodes a 62-kDa ring finger-containing protein that localizes to the cytoplasmic granules in NK cells. Further study of this gene may add to our understanding of cytotoxic processes common to NK cells, CTL, and activated macrophages.

NK cells are lymphocytes that kill tumor or virus-infected cells without prior sensitization (1, 2). However, it is well established that the cytolytic activity of NK cells can be enhanced by a number of cytokines, including IL-2, IFNs (α, β, and γ), IL-12, and IL-15 (3, 4, 5, 6). These cytokines activate NK cells to kill with greater efficiency and broader specificity. IL-2-activated NK cells are responsible for the majority of the lymphokine-activated killing activity observed in vivo (7). High doses of IL-2, with or without administration of lymphokine-activated cells, can induce tumor regression in experimental animals and in some patients (8, 9). However, a number of side effects and toxicities associated with high dose IL-2 therapy have been observed in treated cancer patients that have limited its success and utility (10, 11, 12).

A better understanding of the molecular and cellular events associated with cytokine enhancement of NK function may provide a means for generating and maintaining cells with maximal cytolytic activity and therefore facilitate the use of activated NK cells in cancer, and perhaps also viral, immunotherapy. It has been demonstrated that the augmentation of NK lytic activity by IL-2 or IFN requires de novo RNA and protein synthesis (13, 14). However, it is not clear which gene products are responsible for the increased killing activity and expanded target repertoire. Activated NK cells have increased expression of perforin and granzymes A and B (15, 16). The up-regulation of adhesion molecules CD2, CD11a, and CD54 upon stimulation may also be partially responsible for increased NK lytic activity (17). To further define the events responsible for cytokine augmentation of NK function, we have examined changes in NK gene expression upon IFN-β and IL-2 stimulation.

In the studies presented in this paper the NK clone 3.3 was used as a model to study changes in gene expression in cytokine-activated NK cells. This cloned line is phenotypically and functionally representative of the majority of NK cells in the peripheral blood (i.e., CD16+, CD56+, CD3, CD4, CD8, TCR) (18, 19). Like peripheral blood NK cells, the cytolytic activity of NK3.3 is enhanced after treatment with IL-2 or IFN-β. We have previously found that the protooncogene c-myb is up-regulated by both IL-2 and IFN-β in NK3.3 cells and appears to play an important role in the regulation of NK cytolysis (20). To identify new genes associated with cytokine augmentation of NK function, mRNA isolated from IFN-β-stimulated NK3.3 cells was used to construct a cDNA library. Differential screening was performed to identify a set of novel genes with elevated expression in IFN-β-stimulated NK cells compared with unstimulated cells. Sequence analysis revealed 46 novel genes. This paper describes the characteristics of one of these novel clones, which has been designated NKLAM, for natural killer lytic-associated molecule.3

The derivation and characterization of the human NK clone, NK3.3, have been described in detail previously (18, 20, 21, 22). This IL-2-dependent NK clone is maintained in continuous culture at 2.5–3 × 105 cells/ml in NK medium, composed of RPMI 1640 supplemented with 25 mM HEPES buffer, 2 mM (1%) glutamine, 15% heat-inactivated FBS, and 15% Lymphocult-T as a source of IL-2 (Lymphocult-T, Biotest Diagnostics, Denville, NJ). The NK-sensitive, erythroleukemia cell line K562, the T cell tumor line MOLT-4, the B lymphoblastoid cell line JY, the myeloma cell line ARH77, the breast cancer cell line ZR-75–1, the fibroblast cell line A23, and the neuroglioma line H4 were maintained in complete medium composed of RPMI 1640 supplemented with 25 mM HEPES, 2 mM glutamine, and 10% heat-inactivated FBS. All cell culture reagents, unless specified, were purchased from Life Technologies (Gaithersburg, MD). The fibroblast cell line (A23) was provided by Dr. Sam Goldstein (University of Arkansas for Medical Sciences, Little Rock, AR), and COS-7 cells were purchased from American Type Culture Collection (Manassas, VA). These cell lines were maintained in DMEM supplemented with 10% FBS. Human liver mRNA was isolated from tissue obtained by postmortem autopsy. rIL-2 was a gift from Hoffmann-La Roche (Nutley, NJ), and rIFN-β was a gift from Triton Biosciences (Alameda, CA).

PBL were isolated from healthy donors by centrifugation on Ficoll-Hypaque density gradients. Monocytes were obtained by two rounds of adherence followed by scraping. The purity of this population was between 80–96%. Monocytes were incubated with or without IFN-β (10,000 U/ml) for 24 h. T cells were separated by E-rosetting as previously described (23). SRBC were subsequently lysed from rosetted cells using ACK lysis buffer (150 mM NH4Cl, 1 mM KHCO3, and 0.1 mM Na2EDTA) to obtain enriched T cells. For generation of CTL, CD8+ T cell subpopulations were separated from E rosette-purified T cells using anti-CD8 Ab and anti-mouse IgG magnetic beads (Advanced Magnetics, Cambridge, MA). CD8+ T cells were then incubated with irradiated (5000 rad) allogeneic JY B lymphoblastoid cells in mixed lymphocyte cultures. Equal numbers of CD8+ T cells and irradiated JY cells were cultured in RPMI and 10% FBS at 5 × 105 cells/ml. After 8 days, cells were placed in NK medium. Fourteen days after initiation of the primary MLC, cultures were restimulated with irradiated JY cells. Alloreactive CTL with high cytolytic activity against JY were maintained by continuous growth in NK medium at a concentration of 5 × 105 cells/ml and restimulation with equal numbers of irradiated JY cells every 14 days. In experiments to examine NKLAM RNA expression upon antigenic stimulation, CTL cultures, which had not been restimulated with JY for 14–18 days, were mixed with irradiated JY cells at an E:T cell ratio of 5:1 for 0–4 h. PBL enriched for NK cells were obtained by first removing T cells by E rosetting, followed by the removal of B cells and remaining T cells by immunomagnetic bead separation, using CD19-specific mAb FMC63 and CD3-specific Ab OKT3 followed by goat anti-mouse IgG magnetic beads. T cell-enriched populations were routinely >96% CD3+ by flow cytometric analysis. NK-enriched populations were at least 85% CD16+ and/or CD56+ and <1% CD3+.

NK3.3 cells that had been incubated in medium without IL-2 for 24 h were stimulated for 4 h with 5000 U/ml of IFN-β. Total RNA was then isolated by guanidinium isothiocyanate extraction as previously described (24), and poly(A) mRNA was selected on an oligo(dT)-cellulose column (Boehringer Mannheim, Indianapolis, IN). In the generation of the first library, RNA molecules >500 bases were used as templates for the synthesis of DNA:RNA hybrids using random primers. To generate a second, NKLAM-specific library, synthesis of the first strand of cDNA was performed using a specific primer derived from the 5′ sequence of the original NKLAM cDNA clone (5′-GACGGGCCATATCGCATGTC-3′). Double-stranded DNA was synthesized, and EcoRI linkers were added to the cDNAs and ligated to λZAP II vector arms (Stratagene, La Jolla, CA). The DNA was packaged within phage and used to infect XL1-Blue Escherichia coli cells. In the original cDNA library, differential screening of 106 plaques containing cDNA inserts was performed by transferring plaque DNA onto duplicate nitrocellulose filters. 32P-labeled cDNA synthesized from poly(A) mRNA from IFN-β-treated and IL-2-starved NK3.3 cells were used as probes. The second library was screened using the original NKLAM cDNA as a probe.

Bluescript plasmids containing the cDNA inserts were excised from positive phages by coinfection with helper phage VCSM13 (Stratagene). In some cases, fragments of NKLAM were cloned into M13 to generate ssDNA for sequencing. Dideoxy sequencing using Sequenase or Taq polymerase (U.S. Biochemical Corp., Cleveland, OH) was performed on NKLAM single-stranded and double-stranded cDNA as described by the supplier. Sequencing reactions were initiated using vector-specific or NKLAM-specific oligonucleotide primers. Sequence analysis was performed using the IntelliGenetics PcGene program and GCG Wisconsin package. GenBank, EMBL, PIR, and SwissProt databases were searched using FASTA and BLAST algorithms for homologies (25, 26, 27). Additional programs (Psignal, RaoArgos, TMpred, and PROSITE) were used to predict features of the NKLAM protein.

NK and CTL assays were performed as previously described (18, 20). All conditions were set up in triplicate wells at E:T cell ratios ranging from 30:1 to 2:1. 51Cr release was measured after 4 h of incubation. The percent specific lysis of targets was calculated from each ratio by the formula: % specific lysis = [(experimental release − spontaneous release)/(maximum release − spontaneous release)] × 100. Spontaneous release was measured from target cells incubated alone. Maximum release was determined by adding detergent (0.1 M hexadecyltriammonium bromide) to target cells. The SD of triplicate values in all experiments did not exceed 10%.

Total RNA was isolated using Tri Reagent (Molecular Research Center, Cincinnati, OH), and poly(A)+ RNA was isolated using the RiboSep mRNA Isolation Kit (Collaborative Biomedical Products, Bedford, MA) according to the manufacturers’ protocols. RNA (usually 10 μg/lane) was electrophoresed through 1.5% agarose/6% formaldehyde gels. After capillary transfer of RNA to Magna Charge nylon membranes (Micron Separations, Westboro, MA) and UV cross-linking, blots were prehybridized and then hybridized as previously described (20) using 32P-labeled NKLAM cDNA. All radiolabeled cDNA probes were prepared by random priming (DECA prime II, Ambion, Austin, TX) and had sp. act. of at least 1 × 109 cpm/μg DNA. 32P-radiolabeled dCTP was purchased from Amersham (Arlington Heights, IL). The blots were washed and then exposed to Kodak XAR5 (Eastman Kodak, Rochester, NY) or Fuji (Fuji, Tokyo, Japan) film at −70°C under intensifying screens. Blots for quantitative analysis were exposed to a PhosphorImager screen and then analyzed using a PhosphorImager SF machine (Molecular Dynamics, Sunnyvale, CA). Control probes used in these studies include cDNAs for IFN-γ and ribosomal protein S3 (RPS3). Both were obtained from the American Type Culture Collection. Quantitative analysis of RNA was performed by normalizing hybridization to the levels of RPS3 RNA.

Recombinant NKLAM proteins were generated as fusion proteins containing thioredoxin and a polyhistidine moiety using the pET TRX Fusion System 32 (Novagen, Madison, WI). NKLAM gene sequences encoding aa 31–319 and 272–587 were cut out from cloned cDNA; BamHI linkers (New England Biolabs, Beverly, MA) appropriate for in-frame expression of NKLAM sequences were ligated to the fragments. These products were ligated into the BamHI site of the pET32 vector. Vectors containing fragments in the correct orientation were used to transform expressing E. coli strain BL21 (DE3) pLysS (Novagen). Recombinant proteins were expressed and purified using metal affinity chromatography according to pET TRX Fusion System 32 manufacturer’s recommendation.

Full-length NKLAM was cloned into the eukaryotic expression vector pcDNA3 (Invitrogen, San Diego, CA). This vector was used to generate in vitro translated NKLAM using the TNT-coupled reticulocyte lysate system, according to the manufacturer’s protocol (Promega, Madison, WI). To study NKLAM expression in vivo, COS-7 cells were transfected with pcDNA3 containing NKLAM. Briefly, COS-7 cells were transfected with pcDNA3 or pcDNA3-NKLAM by electroporation using a Life Technologies Cell Porator, followed by selection in G418-containing medium for 48 h. Whole cell extracts from these transfected cells were prepared by cell lysis with 1% SDS and sonication. The extracts were denatured and reduced by boiling in SDS and DTT-containing Laemmli buffer (28) for use in immunoblotting experiments.

Rabbits were immunized with the purified recombinant NKLAM proteins. Sera were prepared, and Abs were affinity purified exactly as described previously (29), except an additional column with immobilized bacterial thioredoxin was used for serum absorption. These Abs to thioredoxin were eluted and used as control sera. Abs to residues 31–319 (N-terminal part of NKLAM protein) were designated AbN; Abs to aa 272–587 (C-terminal part of NKLAM protein) were designated AbC.

Immunoprecipitation was performed as described previously (29). NK3.3 cells were incubated overnight in RPMI 1640 medium containing 10% FBS. Cytoplasmic extracts prepared from these cells are referred to as unstimulated cell extracts. After changing to IL-2-containing NK medium, an aliquot of NK3.3 cells was incubated for an additional 6 h. Cytoplasmic extracts prepared from these cells are referred to as stimulated cell extracts. Cytoplasmic extracts were prepared by Nonidet P-40 lysis. Briefly, cells were washed twice with PBS and resuspended in RIPA buffer (40 mM Tris-Cl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM PMSF, 2 mM benzamide, 50 μg/ml aprotinin, 5 μg/ml leupeptin, and 0.5% Nonidet P-40). The lysate was vortex mixed for 30 s, and the nuclei were spun out by centrifugation for 5 min at 8000 × g in a refrigerated microcentrifuge. The supernatant was designated the cytoplasmic extract and used for subsequent studies. Extracts equivalent to 20 × 106 cells were incubated with 5 μg of the appropriate Abs. The Ag-Ab complexes were precipitated with protein A-agarose (Life Technologies), boiled and reduced in SDS buffer containing DTT, and resolved on polyacrylamide gels using the Laemmli buffer system (28). Prestained m.w. markers (Benchmark prestained protein ladder, lot JJCB15) were obtained from Life Technologies. Protein was transferred to polyvinylidene difluoride membranes (30), probed with primary Abs (0.5 μg/ml), and developed with peroxidase-conjugated secondary Abs and SuperSignal Substrate (Pierce, Rockford, IL).

Phosphorothioate-modified NKLAM sense (S) and antisense (AS) oligodeoxynucleotides (ODN) were generated (Cruachem, Dulles, VA) and used to inhibit NKLAM expression in NK3.3 cells and CTL. In the experiments presented in Figs. 8–10, ODN designated NKLAM AS8, corresponding to nt 694–713 of NKLAM, were used. Additional S and AS constructs, corresponding to nt 684–701, were also employed. These were among several S and AS ODN made and tested; the two AS ODN used in the experiments presented were found to be the most effective in selectively reducing the level of NKLAM mRNA in NK3.3 cells (data not shown). Myc antisense phophorothioate-ODN (designated Myc AS; 5′-AACGTTGAGGGGCAT-3′) were used as a control. This ODN corresponds to the region of c-myc mRNA at the AUG start codon and has been used by others to suppress c-myc expression. ODN were ethanol precipitated twice and resuspended in RPMI medium before use. Before addition of oligonucleotides, NK3.3 cells or CTL were cultured for 24 h in fresh, complete NK medium. The cells were then transferred to medium without IL-2. NKLAM S, AS, and Myc AS ODN were added to the cells at a concentration of 15 μM and introduced into the cells by electroporation. Electroporation was performed in a 0.4-cm gap electroporation chamber in 1 ml of RPMI 1640 medium with 10% FBS at a cell concentration of 4 × 106/ml. For electroporation of NK3.3 cells, the Cell-Porator was set at a capacitance of 800 μF and a voltage of 250. Electroporation settings for CTL were 1180 μF and 280V. After electroporation, cells were allowed to rest in the electroporation chamber for 1 h. Cells were then placed in IL-2-supplemented medium at a concentration of 2 × 105/ml. Two sets of cells were electroporated without ODN; one set was incubated with IL-2-containing medium (E), and the other set was incubated without IL-2 to act as an unstimulated control (−). After incubation for various periods of time, the cells were counted and examined for viability by trypan blue dye exclusion, and tested for cytolytic activity against the NK-sensitive erythroleukemia cell line K562 or the CTL-specific target JY using 4-h 51Cr release assays, then an aliquot was prepared for RNA isolation and characterization.

Cytoplasmic granules were purified from NK3.3 cells according to the procedure described previously (31) with modifications. In brief, cells were collected, washed with ice-cold PBS followed by washing with ice-cold PIPEG buffer (10 mM PIPES (pH 6.8), 4 mM EGTA, 2 mM benzamide, and 8% sucrose), and resuspended in PIPEG buffer supplemented with 1 mM PMSF, 50 μg/ml aprotinin, and 5 μg/ml leupeptin to a final concentration of 5 × 107 cells/ml. The cell suspension was saturated with nitrogen at 400 psi in a cavitation bomb (Parr Instruments, Moline, IL) for 30 min at 4°C, and cells were lysed by decompression. The resulting lysate was adjusted to 8 mM MgCl2 and digested with DNase I for 20 min at room temperature. Nuclei and intact cells were removed by successive filtration through 5- and 3-μm pore size Nucleopore filters (Nucleopore, Pleasanton, CA). The resulting filtrate was cooled on ice and layered on top of a preformed 10–50% gradient of Optiprep (iodixanol) medium (Life Technologies). The gradient contained PIPEG buffer supplemented with 1 mM PMSF, 5 μg/ml aprotinin, and 0.5 μg/ml leupeptin. Since Optiprep medium has a relatively high osmolarity (260 mOsm), sucrose concentrations were varied throughout the length of the gradient to maintain osmolarity between 300–315 mOsm. Cell lysates were fractionated by centrifugation in an angle-type rotor for 2 h at 100,000 × g at 4°C. Fractions were collected from the bottom of the tube, and their densities were determined by weighing on a high precision analytical balance. After boiling in SDS and DTT-containing Laemmli buffer, fractions were separated on a 12% polyacrylamide gel. Western blotting was performed using a mixture of Abs to NKLAM and granzyme B (Biodesign, Kennebunk, ME).

The strategy for cloning cytolysis-specific genes consisted of selecting genes expressed at elevated levels in IFN-β-stimulated NK3.3 cells compared with low or no expression in unstimulated NK3.3 cells. Since NK3.3 is dependent upon IL-2 for growth, IFN-β, rather than IL-2, was chosen as a stimulus to increase the chance of identifying genes important in cytolysis rather than in proliferation. Activation studies are performed by first starving NK3.3 cells of IL-2 for 18–24 h before cytokine stimulation. This treatment has no significant effect on cell viability; however, the killing activity of these cells is diminished by up to 95%. If these cells are then treated with IL-2 or IFN-β, augmentation of NK3.3 lytic activity can be observed as early as 1 h after stimulation (20). Proliferation is seen within 24 h after IL-2 restimulation, but is not induced by IFN-β.

A random-primed library in λ Zap II was constructed using cDNA derived from NK3.3 cells stimulated with 5000 U/ml of IFN-β for 4 h. Plaques (1 × 106) were differentially screened using radiolabeled cDNA derived from poly(A) mRNA from starved and IFN-β-treated NK3.3 cells. Clones were selected that hybridized strongly to RNA from IFN-β-treated NK3.3 cells but weakly or not at all to RNA from nonstimulated NK3.3 cells. After three rounds of screening, 56 plaques were identified. After partial sequencing of these cDNAs, 46 clones encoded genes that appeared to be unique and previously undescribed.

Clones with unique sequences were used as probes to screen RNA from IL-2, IFN-β, and starved NK3.3 cells by Northern blot analysis. One of the unique cDNA clones, designated NKLAM, hybridized to a 2.9-kb RNA transcript that was expressed at a 5- to 8-fold higher level in both IL-2- and IFN-β-stimulated NK3.3 cells compared with unstimulated cells. Since this cDNA clone was not full length, the library was rescreened using this clone as a probe. This resulted in 20 additional overlapping clones, but none contained the complete 5′ end of NKLAM. Therefore, a second NK 3.3 cDNA library was created using NKLAM-specific primers. From this library, three additional overlapping clones containing the missing 5′ end of NKLAM were identified, and the entire gene was sequenced (Fig. 1).

FIGURE 1.

Derived amino acid sequence from NKLAM cDNA. Predicted signal sequence and transmembrane regions are underlined. Sites for potential N-linked glycosylation are indicated by asterisks.

FIGURE 1.

Derived amino acid sequence from NKLAM cDNA. Predicted signal sequence and transmembrane regions are underlined. Sites for potential N-linked glycosylation are indicated by asterisks.

Close modal

The full-length NKLAM cDNA is 2874 nt. It contains a single open reading frame of 1761nt, encoding 587 aa, a 240-bp 5′-untranslated region, and 836-bp 3′-untranslated region, followed by a 37-bp poly(A) tail (Fig. 2,A). The presumed start codon (ATG), located at position 241, has a Kozak consensus sequence critical for initiation of translation (32, 33). Sequence analysis of NKLAM using BLAST and FASTA programs revealed significant homology to three sequences in GenBank. The alignments with NKLAM are shown in Fig. 3; only the regions of significant homology are depicted. The gene with the highest homology to NKLAM encodes a blood-meal-induced protein of 663 aa associated with mosquito reproduction. Another gene with high homology to NKLAM is the C17H11.6 gene product of the nematode Caenorhabditis elegans. Its function is unknown. The BLAST program also identified a murine clone (MGEG-154) with high homology to two stretches of sequence at the 3′ end of NKLAM. This clone, isolated from mouse ovary and overexpressed in mouse testis, has been reported to hybridize strongly to two RNA transcripts of 4.2 and 3 kb (34). It is likely that the portion of MGEG-154 homologous to NKLAM is responsible for hybridization to the 3-kb transcript. The function of this gene is also unknown.

FIGURE 2.

A, Schematic diagram of NKLAM cDNA. It consists of a 240-bp 5′-noncoding region, 1761-bp coding region, and 3′-untranslated region with poly(A) tail. The predicted ATG start codon is shown, compared with the Kozak consensus sequence. B, Schematic diagram of the predicted NKLAM protein. It indicates the position of a potential signal sequence, three predicted transmembrane regions, and two potential sites of N-linked glycosylation.

FIGURE 2.

A, Schematic diagram of NKLAM cDNA. It consists of a 240-bp 5′-noncoding region, 1761-bp coding region, and 3′-untranslated region with poly(A) tail. The predicted ATG start codon is shown, compared with the Kozak consensus sequence. B, Schematic diagram of the predicted NKLAM protein. It indicates the position of a potential signal sequence, three predicted transmembrane regions, and two potential sites of N-linked glycosylation.

Close modal
FIGURE 3.

Alignment of the predicted NKLAM protein with similar areas in homologous proteins in mosquito, C. elegans, and the mouse. Accession numbers are I64695, 1813871, 1707258, and 1730145, respectively. Numbers in columns indicate the position of the amino acids in each of the proteins. Sequences in boldface denote the regions of greatest homology.

FIGURE 3.

Alignment of the predicted NKLAM protein with similar areas in homologous proteins in mosquito, C. elegans, and the mouse. Accession numbers are I64695, 1813871, 1707258, and 1730145, respectively. Numbers in columns indicate the position of the amino acids in each of the proteins. Sequences in boldface denote the regions of greatest homology.

Close modal

NKLAM contains a potential signal sequence, with a predicted cleavage site between aa 45 and 46 identified on the basis of the (−3, −1) rule (35). Computer analysis of the NKLAM amino acid sequence performed with the RaoArgos and TMpred programs detects three well-defined transmembrane segments located between aa 358–380, 409–430, and 513–531, therefore identifying a strongly preferred model of NKLAM as a transmembrane protein with a predicted size of 62 kDa (Fig. 2 B) (36, 37). Sequences corresponding to the first two putative transmembrane regions are highly homologous among NKLAM, the mosquito protein, the C. elegans sequence, and GEG-154. Using the PROSITE patterns database, several potential sites for protein modification were identified in NKLAM (38). These include sites for N-linked glycosylation at aa 248 and 562, and several potential amidation and N-myristylation sites throughout the molecule. There is also a potential cAMP- and cGMP-dependent protein kinase phosphorylation site, three potential protein kinase C phosphorylation sites, and two potential casein kinase II phosphorylation sites in the putative nontransmembrane domains of the NKLAM protein.

The predicted N-terminal region of NKLAM contains 26 cysteines within a 213-aa region from 119 to 332, which fall into three cysteine-rich clusters. Sequence analysis of these clusters revealed two potential zinc binding domains located between aa 115–161 and 282–333. These two domains match the consensus sequence of the C3HC4 ring finger family (C-X2-C-X(9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27)-C-X(1, 2, 3)-H-X(2, 3)-C-X2-C-X4–48(4–48)-C-X2-C, where C and H denote cysteine and histidine, and X represents other amino acids) (39, 40, 41). An additional cysteine-rich domain located between aa 202 and 252 does not perfectly match other well-defined cysteine-rich clusters and may form a novel zinc finger domain. The three cysteine-rich domains appear to be highly conserved, in that they are also found in the homologous mosquito, C. elegans, and mouse GEG-154 sequences (Fig. 3). BLAST sequence analysis of this cysteine-rich region of NKLAM revealed additional homology to other proteins with ring finger domains. The best similarity was seen to KIAA1, a predicted protein from a cDNA clone isolated from human KG-1 myeloblast cells (accession no. D79983) and the ARI ring finger protein found in the central nervous system of Drosophila melanogaster (accession no. X98309). The location, distribution, and orientation of these three cysteine-rich domains in the three proteins are also similar, suggesting a functional relationship.

The kinetics of induction of NKLAM mRNA expression and corresponding cytolytic function of NK3.3 cells after stimulation with IL-2 and IFN-β were assessed. In these experiments, NK3.3 cells were first cultured in IL-2-free medium for 18–24 h (starved) and then stimulated with IL-2 (200 U/ml) or IFN-β (10,000 U/ml) for the indicated times. Cells were harvested for RNA isolation and Northern blot analysis; the cytolytic activity of NK3.3 cells using the NK-sensitive target K562 was also monitored concurrently in 51Cr release assays. As shown in Fig. 4, A and B, NKLAM RNA expression increases within 2 h of IFN stimulation, peaks at 4–6 h (6-fold increase), and then declines. This closely parallels the kinetics of IFN-mediated augmentation of killing (Fig. 4,C). When cells are stimulated with IL-2, NKLAM RNA levels also increase within the first hour and rise 5- to 6-fold over a 6- to 8-h period. However, unlike IFN stimulation, where levels then drop, with IL-2 stimulation, NKLAM levels continue to rise, although much more slowly, throughout the 6- to 12-h period (Fig. 5, A and B). This pattern of kinetics exactly parallels the kinetics of IL-2 augmentation of NK3.3 cytolytic activity (Fig. 5 C). Therefore, although the patterns of IL-2- and IFN-mediated augmentation of killing and NKLAM expression are different for each cytokine, the kinetics of cytokine induction of NKLAM and cytolysis are identical.

FIGURE 4.

Correlation between IFN-β-stimulated cytolytic activity and IFN-β-stimulated induction of NKLAM mRNA in NK3.3 cells. IL-2-starved, NK3.3 cells were treated with IFN-β (10,000 U/ml) for 0–12 h. At the indicated time points, an aliquot of cells was assessed for NK cytolytic activity against 51Cr-labeled K562 cells. In parallel, a second aliquot of cells was used to prepare RNA for Northern blot analysis. A, Northern blot analysis indicating the kinetics of IFN-β induction of NKLAM mRNA in NK3.3 cells. Blots were probed with a 32P-labeled NKLAM cDNA probe (a 2-kb fragment corresponding to the entire 3′ end of NKLAM), an IFN-γ cDNA probe (ATCC no. 39042), and an RPS3 cDNA probe (ATCC no. 65825). B, Quantitative analysis of kinetics of IFN-β induction of NKLAM mRNA. The level of mRNA at each time point was calculated as a function of the volume of radioactivity in hybridizing bands. RNA values for NKLAM and IFN-γ at each time point were normalized to the levels of the noninducible, highly stable RPS3 transcripts. The relative amount of NKLAM mRNA at each point was determined by setting the level of NKLAM mRNA in IL-2-starved, unstimulated NK3.3 cells (time zero) to a value of 1. Quantitative analysis was performed using a PhosphorImager. C, Kinetics of IFN-β stimulation of cytolytic activity in NK3.3 cells. At each of the designated time points, an aliquot of NK3.3 cells was removed and assayed for cytolytic activity in 4-h 51Cr release assays using radiolabeled K562 cells as targets. Results are expressed as the percent specific lysis at an E:T cell ratio of 20:1.

FIGURE 4.

Correlation between IFN-β-stimulated cytolytic activity and IFN-β-stimulated induction of NKLAM mRNA in NK3.3 cells. IL-2-starved, NK3.3 cells were treated with IFN-β (10,000 U/ml) for 0–12 h. At the indicated time points, an aliquot of cells was assessed for NK cytolytic activity against 51Cr-labeled K562 cells. In parallel, a second aliquot of cells was used to prepare RNA for Northern blot analysis. A, Northern blot analysis indicating the kinetics of IFN-β induction of NKLAM mRNA in NK3.3 cells. Blots were probed with a 32P-labeled NKLAM cDNA probe (a 2-kb fragment corresponding to the entire 3′ end of NKLAM), an IFN-γ cDNA probe (ATCC no. 39042), and an RPS3 cDNA probe (ATCC no. 65825). B, Quantitative analysis of kinetics of IFN-β induction of NKLAM mRNA. The level of mRNA at each time point was calculated as a function of the volume of radioactivity in hybridizing bands. RNA values for NKLAM and IFN-γ at each time point were normalized to the levels of the noninducible, highly stable RPS3 transcripts. The relative amount of NKLAM mRNA at each point was determined by setting the level of NKLAM mRNA in IL-2-starved, unstimulated NK3.3 cells (time zero) to a value of 1. Quantitative analysis was performed using a PhosphorImager. C, Kinetics of IFN-β stimulation of cytolytic activity in NK3.3 cells. At each of the designated time points, an aliquot of NK3.3 cells was removed and assayed for cytolytic activity in 4-h 51Cr release assays using radiolabeled K562 cells as targets. Results are expressed as the percent specific lysis at an E:T cell ratio of 20:1.

Close modal
FIGURE 5.

Correlation between IL-2-stimulated cytolytic activity and IL-2-stimulated induction of NKLAM mRNA in NK3.3 cells. A, Northern blot analysis of IL-2 induction of NKLAM mRNA in NK3.3 cells. Cells were stimulated with 200 U/ml of IL-2 for the times indicated and treated as described in Fig. 4,A. B, Quantitative analysis of the kinetics of IL-2 induction of NKLAM mRNA was performed as described in Fig. 4,B. C, Kinetics of IL-2 stimulation of cytolytic activity in NK3.3 cells. Lysis of K562 at an E:T cell ratio of 20:1 was assessed as described in Fig. 4 C.

FIGURE 5.

Correlation between IL-2-stimulated cytolytic activity and IL-2-stimulated induction of NKLAM mRNA in NK3.3 cells. A, Northern blot analysis of IL-2 induction of NKLAM mRNA in NK3.3 cells. Cells were stimulated with 200 U/ml of IL-2 for the times indicated and treated as described in Fig. 4,A. B, Quantitative analysis of the kinetics of IL-2 induction of NKLAM mRNA was performed as described in Fig. 4,B. C, Kinetics of IL-2 stimulation of cytolytic activity in NK3.3 cells. Lysis of K562 at an E:T cell ratio of 20:1 was assessed as described in Fig. 4 C.

Close modal

The levels of two other species of RNA were measured in parallel. RPS3 encodes the ribosomal structural protein S3 (42). This RNA has a very long half-life, is extremely stable and constant throughout the cell cycle, and therefore acts as a good control of total RNA levels. Its expression is not significantly altered by cytokines or cell activation. Levels of RNA were normalized using the RPS3 probe in the quantitation of NKLAM. The other RNA that was measured encodes IFN-γ; it was used to monitor NK cell activation. Ye et al. (43) reported that NK3.3 cells produce IFN-γ after IL-2 treatment. Here, we show that both IL-2 and IFN-β induce the expression of IFN-γ mRNA in NK3.3 cells. (Figs. 4 and 5) The kinetics of induction of IFN-γ mRNA in NK3.3 cells upon IL-2 and IFN-β stimulation closely parallel the kinetics of induction of NKLAM. This suggests that there may be a common regulatory pathway for IFN-γ and NKLAM RNA induction in NK cells.

To determine whether the increase in NKLAM mRNA in cytokine-stimulated NK3.3 cells was accompanied by a commensurate increase in NKLAM protein production, we employed a combination of immunoprecipitation and Western blot analysis using polyclonal Abs to NKLAM raised in rabbits. Extracts from unstimulated and IL-2-stimulated NK3.3 cells were immunoprecipitated with Abs raised against the N-terminal part of NKLAM (AbN, see Fig. 6, lanes 2 and 3) and Abs to thioredoxin (Fig. 6, lane 1). The precipitated material was analyzed by Western blot with Abs to the C-terminal part of NKLAM (AbC). As a positive control, the in vitro translation product of NKLAM was run on SDS-PAGE and immunoblotted with AbC (Fig. 6, lane 4). Additional controls include whole cell lysates from COS-7 cells transiently transfected with the NKLAM-pcDNA3 expression vector or empty vector (lanes 5 and 6, respectively). Fig. 6 shows that several bands could be detected in IL-2-stimulated NK3.3 cells after immunoprecipitation with NKLAM-specific AbN (lane 3). These bands were absent from extracts precipitated with nonrelated Abs to thioredoxin and were barely detectable in extracts from unstimulated NK3.3 cells precipitated with AbN (Fig. 6, lanes1 and 2, respectively). The apparent molecular masses of the two lower bands seen in IL-2-stimulated NK3.3 cells are approximately 65 and 75 kDa. The 65-kDa band migrates in a fashion consistent with the molecular mass of NKLAM determined by computer analysis, in vitro translation, and transient transfection experiments, as shown. The nature of the other bands is under investigation.

FIGURE 6.

Immunodetection of NKLAM protein in NK3.3 cells. Extracts from IL-2-stimulated (lanes 1 and 3) or unstimulated (lane 2) NK3.3 cells were precipitated with anti-thioredoxin Abs (lane 1), or NKLAM-specific AbN (lanes 2 and 3). Precipitated material was analyzed by Western blotting using NKLAM-specific AbC. Controls are the in vitro translation product of NKLAM (lane 4) and extracts from COS-7 cells transiently transfected with pcDNA3 containing NKLAM (lane 5) or empty vector (lane 6). After SDS-PAGE, these samples were immunoblotted using NKLAM AbC. The molecular weights of the prestained protein ladder (Benchmark, Life Technologies) are displayed on the left of the figure.

FIGURE 6.

Immunodetection of NKLAM protein in NK3.3 cells. Extracts from IL-2-stimulated (lanes 1 and 3) or unstimulated (lane 2) NK3.3 cells were precipitated with anti-thioredoxin Abs (lane 1), or NKLAM-specific AbN (lanes 2 and 3). Precipitated material was analyzed by Western blotting using NKLAM-specific AbC. Controls are the in vitro translation product of NKLAM (lane 4) and extracts from COS-7 cells transiently transfected with pcDNA3 containing NKLAM (lane 5) or empty vector (lane 6). After SDS-PAGE, these samples were immunoblotted using NKLAM AbC. The molecular weights of the prestained protein ladder (Benchmark, Life Technologies) are displayed on the left of the figure.

Close modal

PBMC were fractionated into monocytes, T cells, and NK cells, and each subpopulation was examined for the presence of NKLAM mRNA. The results are presented in Fig. 7 A. In this graphic representation, the baseline, low level of NKLAM in unstimulated, starved NK3.3 cells was set to a value of 1 (bar 10); 4-h IFN-β-stimulated NK3.3 cells (bar 11) have 6 times more NKLAM RNA than the unstimulated cells. Peripheral blood monocytes (bar 1) constitutively express high levels of NKLAM compared with unstimulated NK3.3 cells. IFN-β stimulation significantly enhances their expression of NKLAM (bar 2). Unstimulated peripheral blood NK cells also express significant levels of NKLAM (equivalent to that of IFN-β-stimulated NK 3.3 cells; bar 3). These levels are increased 3- to 5-fold further by 24-h IFN-β treatment (bar 4). In contrast, peripheral blood T cells express barely detectable levels of NKLAM RNA (bar 7). Neither IL-2 nor PHA stimulation significantly induced NKLAM expression in these purified T cells (bars 8 and 9). However, a CD8+ alloreactive, cytotoxic T cell line generated from these purified T cells by multiple stimulation with irradiated, allogeneic B lymphoblastoid JY cells expressed significant levels of NKLAM RNA (comparable to unstimulated NK3.3 cells; bar 5). Upon stimulation of these CTL with their specific alloreactive target for 4 h, NKLAM RNA was strongly induced, increasing approximately 10-fold in expression compared with unstimulated CTL (bar 6). These results indicate that NKLAM expression is restricted to cells with cytolytic activity and further induced when cytolysis is triggered or enhanced.

FIGURE 7.

Expression of NKLAM mRNA in human PBMC subpopulations and other cell types. Graphs represent the quantitative analysis of hybridization of a 32P-labeled NKLAM cDNA probe to RNA from isolated subpopulations of peripheral blood cells (A) and different cell lines and tissues (B). The relative amount of NKLAM mRNA in each cell type was determined by setting the level of NKLAM mRNA in unstimulated NK3.3 cells (A, bar 10) to a value of 1 after normalizing RNA levels using the RPS3 probe. Quantitative analysis was performed using a PhosphorImager. A, Peripheral blood monocytes, NK cells, T cells, and CTL were obtained as described in Materials and Methods. By flow cytometric analysis, monocytes were 80–96% pure, T cells were >96% CD3+, NK cells were at least 85% CD16+ and/or CD56+ and <1% CD3+, and the CTL line was >99% CD8+. CTL were stimulated with irradiated JY cells at an E:T cell ratio of 5:1 for 4 h. Peripheral blood NK cells, monocytes, and NK3.3 were cultured in complete medium with or without 10,000 U/ml IFN-β for the times indicated. T cells were cultured for 24 or 48 h in complete medium with or without IL-2 (200 U/ml) or PHA (50 μg/ml). B, Cell lines were incubated with or without IFN-β (10,000 U/ml) for 4 h. These included the breast cancer cell line ZR-75–1, the T cell line MOLT-4, the erythroleukemia cell line K562, and the myeloma cell line ARH77. Other cells tested were the neuroglioma cell line H4, the fibroblast line A23, and normal liver obtained from autopsy.

FIGURE 7.

Expression of NKLAM mRNA in human PBMC subpopulations and other cell types. Graphs represent the quantitative analysis of hybridization of a 32P-labeled NKLAM cDNA probe to RNA from isolated subpopulations of peripheral blood cells (A) and different cell lines and tissues (B). The relative amount of NKLAM mRNA in each cell type was determined by setting the level of NKLAM mRNA in unstimulated NK3.3 cells (A, bar 10) to a value of 1 after normalizing RNA levels using the RPS3 probe. Quantitative analysis was performed using a PhosphorImager. A, Peripheral blood monocytes, NK cells, T cells, and CTL were obtained as described in Materials and Methods. By flow cytometric analysis, monocytes were 80–96% pure, T cells were >96% CD3+, NK cells were at least 85% CD16+ and/or CD56+ and <1% CD3+, and the CTL line was >99% CD8+. CTL were stimulated with irradiated JY cells at an E:T cell ratio of 5:1 for 4 h. Peripheral blood NK cells, monocytes, and NK3.3 were cultured in complete medium with or without 10,000 U/ml IFN-β for the times indicated. T cells were cultured for 24 or 48 h in complete medium with or without IL-2 (200 U/ml) or PHA (50 μg/ml). B, Cell lines were incubated with or without IFN-β (10,000 U/ml) for 4 h. These included the breast cancer cell line ZR-75–1, the T cell line MOLT-4, the erythroleukemia cell line K562, and the myeloma cell line ARH77. Other cells tested were the neuroglioma cell line H4, the fibroblast line A23, and normal liver obtained from autopsy.

Close modal

To determine whether expression of the NKLAM transcript was limited to hemopoietic cells, a number of other cell types were screened (Fig. 7 B). NKLAM was not strongly expressed in any of these cell types tested, including fibroblasts, breast cancer cells (ZR-75–1), T cell tumor lines (MOLT-4), and myeloma cell lines (ARH77). There was a low level of expression in liver and in the neuroglioma cell line (H4; bars 3 and 4). However, NKLAM was not induced in any of these cells following a 4-h stimulation with 10,000 U/ml IFN-β. Further analysis of NKLAM expression in a variety of human tissues using a commercially obtained tissue blot (Clontech multiple tissue Northern blot) indicated that NKLAM RNA levels lower than those in liver were seen in pancreas, muscle, and brain (data not shown).

To attempt to establish a link between NKLAM expression and cytolytic function, a series of antisense oligonucleotide experiments were performed. NK3.3 cells were cultured for 24 h in IL-2-supplemented medium and then transferred to complete medium without IL-2. NKLAM S and AS phosphorothioate-modified ODN were introduced into NK3.3 cells by electroporation. As an additional control, a well-defined c-Myc antisense phosphorothioate ODN was used. This ODN corresponds to the region of c-myc mRNA at the AUG start codon and has been used successfully to suppress c-myc expression (44). All ODN were used at a final concentration of 15 μM. After electroporation without (E) or with ODN, cells were placed in IL-2-supplemented medium. One set of electroporated cells were incubated in medium without IL-2 to act as an unstimulated control (−). The results of cytotoxicity experiments using two representative NKLAM antisense ODN (designated AS and AS8) are presented in Fig. 8. NKLAM antisense treatment of NK3.3 reduced its cytotoxic activity against K562 by about 75%, while NKLAM S and Myc AS treatments had minimal effects. These data represent one of four experiments performed with identical results. Fig. 9 demonstrates the specificity and selectivity of the NKLAM antisense ODN for NKLAM. As shown in Fig. 9, A and B, NKLAM antisense treatment of NK3.3 dramatically reduced the levels of NKLAM mRNA (65% decrease) without significantly affecting RPS3 or IFN-γ mRNA expression. This illustrates the specificity of the interaction of the NKLAM AS ODN with NKLAM. Treatment of NK3.3 with Myc AS ODN only slightly reduced the levels of NKLAM mRNA and had no effect on RPS3 or IFN-γ mRNA expression.

FIGURE 8.

NKLAM antisense oligonucleotide-mediated decrease in NK cytolytic activity. NK3.3 cells were transferred to complete medium without IL-2. NKLAM S, AS, and myc AS ODN were added to the medium at concentrations of 15 μM and introduced into the cells by electroporation, as described in Materials and Methods. After the cells rested in the electroporation chambers for 1 h, they were placed in medium supplemented with 200 U/ml IL-2. Control cells were electroporated without ODN. One set of control cells was placed in IL-2 medium to act as an electroporation control (E), and the other set was placed in medium without IL-2 to act as an unstimulated control (−). After incubation for 18 h, 51Cr release assays were performed as described using radiolabeled K562 target cells. Results are expressed as the percent specific lysis of K562. This represents one of four experiments with similar results. SDs of triplicate values in each experiment did not exceed 8%.

FIGURE 8.

NKLAM antisense oligonucleotide-mediated decrease in NK cytolytic activity. NK3.3 cells were transferred to complete medium without IL-2. NKLAM S, AS, and myc AS ODN were added to the medium at concentrations of 15 μM and introduced into the cells by electroporation, as described in Materials and Methods. After the cells rested in the electroporation chambers for 1 h, they were placed in medium supplemented with 200 U/ml IL-2. Control cells were electroporated without ODN. One set of control cells was placed in IL-2 medium to act as an electroporation control (E), and the other set was placed in medium without IL-2 to act as an unstimulated control (−). After incubation for 18 h, 51Cr release assays were performed as described using radiolabeled K562 target cells. Results are expressed as the percent specific lysis of K562. This represents one of four experiments with similar results. SDs of triplicate values in each experiment did not exceed 8%.

Close modal
FIGURE 9.

NKLAM antisense oligonucleotide treatment selectively decreases the expression of NKLAM RNA by NK3.3, but does not affect its IL-2-dependent proliferation. After electroporation without addition of oligonucleotides (E), with 15 μM Myc AS ODN, or NKLAM AS8 ODN (as described above and in Materials and Methods), NK3.3 cells were adjusted to a concentration of 3–5 × 105 cells/ml in IL-2 containing medium. After incubation for 6 h (A and B), RNA was isolated from the cells, electrophoresed, blotted, and hybridized to NKLAM, IFN-γ, and RPS3 cDNA probes. In C, cells were initially seeded at a concentration of 2.3 × 105 cells/ml (relative cell number × 10−3), maintained in culture for 48 h and monitored for growth and viability. A, Quantitative analysis of NKLAM mRNA levels from untreated, electroporated (E), NKLAM AS8, and Myc AS ODN-treated NK3.3 cells, as shown in B. Results are expressed as a percentage of NKLAM mRNA, with levels in the electroporated, untreated control cells set to 100%. B, Northern blot hybridization of RNA isolated from untreated and ODN-treated cells. Hybridization to NKLAM, IFN-γ, and RPS3 cDNA probes is shown. C, Growth curves of control (E), Myc AS-treated, and NKLAM AS8-treated NK3.3 cells. Cells were initially treated as described above; at the times indicated, they were counted, and viability was assessed by trypan blue dye exclusion. Relative cell numbers reflect the numbers of viable cells (×10−3) in each culture; viability was >90% under all conditions except with c-Myc antisense treatment, where viability ranged from 70–85%. Error bars represent the SDs from three independent experiments.

FIGURE 9.

NKLAM antisense oligonucleotide treatment selectively decreases the expression of NKLAM RNA by NK3.3, but does not affect its IL-2-dependent proliferation. After electroporation without addition of oligonucleotides (E), with 15 μM Myc AS ODN, or NKLAM AS8 ODN (as described above and in Materials and Methods), NK3.3 cells were adjusted to a concentration of 3–5 × 105 cells/ml in IL-2 containing medium. After incubation for 6 h (A and B), RNA was isolated from the cells, electrophoresed, blotted, and hybridized to NKLAM, IFN-γ, and RPS3 cDNA probes. In C, cells were initially seeded at a concentration of 2.3 × 105 cells/ml (relative cell number × 10−3), maintained in culture for 48 h and monitored for growth and viability. A, Quantitative analysis of NKLAM mRNA levels from untreated, electroporated (E), NKLAM AS8, and Myc AS ODN-treated NK3.3 cells, as shown in B. Results are expressed as a percentage of NKLAM mRNA, with levels in the electroporated, untreated control cells set to 100%. B, Northern blot hybridization of RNA isolated from untreated and ODN-treated cells. Hybridization to NKLAM, IFN-γ, and RPS3 cDNA probes is shown. C, Growth curves of control (E), Myc AS-treated, and NKLAM AS8-treated NK3.3 cells. Cells were initially treated as described above; at the times indicated, they were counted, and viability was assessed by trypan blue dye exclusion. Relative cell numbers reflect the numbers of viable cells (×10−3) in each culture; viability was >90% under all conditions except with c-Myc antisense treatment, where viability ranged from 70–85%. Error bars represent the SDs from three independent experiments.

Close modal

Fig. 9 C shows the growth curves of NK3.3 after electroporation with 15 μM NKLAM AS ODN or Myc AS ODN compared with those of control, electroporated cells growing in the absence of ODN. The IL-2-dependent proliferation of NK3.3 was unaffected by the NKLAM AS ODN, confirming the selective role of NKLAM in cytokine-enhanced NK cytotoxic activity. As expected, the Myc AS ODN treatment induced growth arrest in NK3.3 cells during the first 24 h of culture. Cumulatively, these data, demonstrating a selective and dramatic reduction in NKLAM mRNA levels and cytotoxic activity in IL-2-activated NK3.3 cells treated with NKLAM AS ODN, with no effect on other gene expression or cell proliferation, strongly support the role of NKLAM in cytokine-activated NK cytolysis.

NKLAM antisense ODN experiments using CTL were performed with similar results. CTL were electroporated without oligonucleotides (E) or with NKLAM S, NKLAM AS, or Myc AS ODN. After incubation in IL-2-containing medium for 24 h, cytolysis against the specific target JY was assessed (Fig. 10,A). NKLAM AS ODN treatment resulted in a 60% decrease in CTL activity. A corresponding, selective decrease in NKLAM RNA levels was seen (Fig. 10 B). NKLAM S and Myc AS ODN had minimal effects. These results suggest a common role for NKLAM in both NK- and T cell-mediated cytotoxicity.

FIGURE 10.

NKLAM antisense oligonucleotide treatment selectively decreases the cytolytic activity of CTL and its NKLAM RNA expression. After electroporation of CTL without ODN (E), with 15 μM NKLAM S ODN, NKLAM AS8 ODN, or Myc AS ODN, cells were incubated in IL-2-containing medium for 6 or 24 h. A, Cytotoxic activity of CTL against the specific target JY was assessed by 4-h 51Cr release assays 24 h after electroporation. Results are expressed as the percent specific lysis of JY at four E:T cell ratios. The SD of triplicate values at each point did not exceed 6%. This killing assay is representative of four separate experiments with similar results. B, Northern blot analysis of RNA from CTL electroporated with nothing (E), NKLAM S, NKLAM AS8, and Myc AS ODN. RNA was prepared from CTL 6 h after treatment with ODN. Hybridization to NKLAM and RPS3 probes is shown.

FIGURE 10.

NKLAM antisense oligonucleotide treatment selectively decreases the cytolytic activity of CTL and its NKLAM RNA expression. After electroporation of CTL without ODN (E), with 15 μM NKLAM S ODN, NKLAM AS8 ODN, or Myc AS ODN, cells were incubated in IL-2-containing medium for 6 or 24 h. A, Cytotoxic activity of CTL against the specific target JY was assessed by 4-h 51Cr release assays 24 h after electroporation. Results are expressed as the percent specific lysis of JY at four E:T cell ratios. The SD of triplicate values at each point did not exceed 6%. This killing assay is representative of four separate experiments with similar results. B, Northern blot analysis of RNA from CTL electroporated with nothing (E), NKLAM S, NKLAM AS8, and Myc AS ODN. RNA was prepared from CTL 6 h after treatment with ODN. Hybridization to NKLAM and RPS3 probes is shown.

Close modal

Studies were performed to determine where NKLAM protein resides in cytolytic cells. Cellular organelles from IL-2-stimulated NK3.3 cells disrupted by nitrogen cavitation were separated by iodixanol (Optiprep) density gradient centrifugation. As shown in Fig. 11, both granzyme B and NKLAM were detected in fractions with densities ranging from 1.06–1.1 g/cm3. These fractions also contained the lysosomal enzyme β-glucuronidase (not shown). The peak expression of both granzyme B and NKLAM comigrated at a density of 1.087, which is consistent with published reports of the density of cytolytic granules from CTL and NK cells (31, 45). The peaks of both granzyme B and NKLAM were well separated from the mitochondrial fractions, as assessed by succinate dehydrogenase activity (data not shown). This enzyme was chosen because of all cellular organelles, mitochondria are closest in density to cytoplasmic granules. These results show that NKLAM colocalizes with granzyme B, a known component of cytolytic granules, at a density characteristic of granules, making it likely that NKLAM resides in the granules in NK cells.

FIGURE 11.

NKLAM protein colocalizes with granzyme B to the cytoplasmic granules of IL-2-stimulated NK3.3 cells. Consecutive fractions from NK3.3 lysates centrifuged through a preformed 10–50% Optiprep gradient were collected and separated on a 12% polyacrylamide gel. Western blotting was performed using a mixture of Abs to granzyme B and NKLAM. The positions of the m.w. markers (Benchmark, Life Technologies) are shown on the right. The positions of granzyme B (Gr) and NKLAM proteins are indicated on the left. The fractions presented range in density from 1.013–1.139 g/cm3. NKLAM and granzyme B are detectable in fractions with densities between 1.06 and 1.1 g/cm3. The peak of both granzyme B and NKLAM expression is in the fraction with a density of 1.087 g/cm3.

FIGURE 11.

NKLAM protein colocalizes with granzyme B to the cytoplasmic granules of IL-2-stimulated NK3.3 cells. Consecutive fractions from NK3.3 lysates centrifuged through a preformed 10–50% Optiprep gradient were collected and separated on a 12% polyacrylamide gel. Western blotting was performed using a mixture of Abs to granzyme B and NKLAM. The positions of the m.w. markers (Benchmark, Life Technologies) are shown on the right. The positions of granzyme B (Gr) and NKLAM proteins are indicated on the left. The fractions presented range in density from 1.013–1.139 g/cm3. NKLAM and granzyme B are detectable in fractions with densities between 1.06 and 1.1 g/cm3. The peak of both granzyme B and NKLAM expression is in the fraction with a density of 1.087 g/cm3.

Close modal

This report describes the initial characterization of a novel gene, designated NKLAM. This gene, isolated from a cDNA library made from IFN-β-stimulated NK3.3 cells, is highly expressed in activated cells with cytotoxic potential, including NK cells, CTL, and macrophages. Kinetic studies demonstrate that NKLAM RNA expression just precedes and strongly correlates with cytokine-enhanced NK cytolytic activity. The full-length NKLAM cDNA clone is 2874nt. The 5′-untranslated region is very GC rich (80% GC) and can potentially form strong secondary structures. This type of sequence is characteristic of genes for growth factors, cytokine receptors, and protooncogenes and is associated with poorly translated transcripts (31, 46).

The single open reading frame encodes 587 aa , beginning with an AUG codon surrounded by sequences that conform to a Kozak consensus sequence for translation initiation. Amino acid sequence analysis predicts the presence of an N-terminal signal sequence with cleavage site located between positions 45 and 46. This cleavage would result in a mature, unprocessed protein of 58 kDa. NKLAM is predicted to have three transmembrane helices crossing the lipid bilayer. The N-terminal region contains three cysteine-rich clusters. Two of these clusters correspond to ring finger motifs; the third may represent a new class of zinc finger domain. Up to 80 proteins have been shown to contain ring finger motifs (39, 41). As is the case with NKLAM, the ring finger tends to reside near the N-termini of these proteins. The precise role of the ring finger domain remains to be elucidated. In some proteins, the ring finger is part of a DNA binding domain. However, other ring finger-containing proteins are part of multiprotein complexes, suggesting that the ring finger domain may also be involved in protein-protein interactions (41). The high degree of homology between NKLAM and a gene product of the nematode C. elegans and of the mosquito, Aedes aegypti, especially in the cysteine-rich and putative transmembrane domains, indicates the potential importance of these molecules, but provides few clues to their function.

Expression of NKLAM is strongly correlated with cytokine augmentation of NK cytolytic function. Both IL-2 and IFN-β enhance cytolytic activity and NKLAM RNA and protein expression in NK3.3 cells and peripheral blood NK cells. In NK3.3, elevated NKLAM RNA is detectable within 1 h of stimulation with either IL-2 or IFN and peaks at about 6 h. This precisely parallels the cytokine-induced augmentation of cytolytic activity. After 6 h, NKLAM levels and cytolytic activity drop in IFN-stimulated cells, while they persist, and even continue to slowly rise, in IL-2-stimulated cells. Therefore, there is direct correlation between the kinetics of cytokine induction of NKLAM and cytolysis.

Cytokine treatment often results in elevation of RNA levels by increasing message stability. For example, IL-2 dramatically enhances the half-life of IFN-γ mRNA in NK3.3 cells, resulting in a significant accumulation and corresponding increase in IFN-γ mRNA (43). However, the half-life of NKLAM RNA, which is approximately 2.5 h in untreated NK cells, is identical in IL-2- and IFN-β-treated cells (data not shown). Therefore, the increased steady state levels of NKLAM mRNA in response to IL-2 and IFN-β are most likely due to cytokine-mediated transcriptional activation of NKLAM.

Data obtained with a combination of immunoprecipitation and Western blot techniques suggest that NKLAM protein is also overexpressed in cytokine-stimulated cells. The set of proteins in IL-2-stimulated NK3.3 cells detected by Western blotting using polyclonal Abs to NKLAM probably corresponds to NKLAM by the following criteria. First, these proteins are not recognized or precipitated by control Abs to thioredoxin purified from the same rabbit antisera. However, the same pattern of proteins is precipitated by Abs to the N- and C-terminal part of NKLAM and is recognized by both these Abs in immunoblotting (data not shown). Second, the 65-kDa protein recognized by NKLAM-specific Abs has an apparent molecular mass consistent with that predicted from the cDNA sequence of NKLAM and is the same size as the in vitro translation product and protein derived by transient transfection of COS-7 cells with NKLAM. The 75-kDa protein may be a product of post-translational modification. The nature of the large protein species migrating in the 110–130 kDa range in these Western blots is still uncertain. The size of the band suggests that it might represent a homodimer of NKLAM or a multiprotein complex. However, this seems unlikely, since all samples were boiled in SDS under reducing conditions before electrophoresis. Alternatively, this band may be another novel protein with homology to NKLAM that reacts with Abs directed against both the N- and C-terminal regions of NKLAM. However, if it is another protein, it is also found in NK cytoplasmic granules, comigrating with NKLAM and granzyme B. Another possibility is that this band represents the protein product of alternatively spliced NKLAM mRNA. This interpretation is consistent with our recent finding of a homologous murine NKLAM transcript with a significantly longer open reading frame than that of human NKLAM. These possibilities are currently being tested.

To further assess the potential functional involvement of NKLAM in the cytolytic process, CTL were stimulated with their specific alloreactive target, JY. NKLAM RNA levels increased >10-fold in CTL after 4 h of target stimulation. Under similar conditions of target stimulation using NK cells as effectors, Salcedo et al. saw a similar increase in granzyme B mRNA accumulation (16). These results lend additional support for the role of NKLAM in cytolysis.

To more directly establish a link between NKLAM expression and cytolytic function, a series of antisense oligonucleotide experiments was performed. Antisense approaches have been successfully used to delineate functional roles for a number of gene products involved in cell-mediated cytolysis, including the involvement of granzymes A, B, and NK-TR in NK lytic activity (47, 48, 49, 50). We have tried to control for potential nonspecific, nonantisense-mediated effects in a variety of ways. The NKLAM antisense constructs used in these experiments do not contain a dG quartet, which has been associated with nonspecific binding of oligomer to proteins (51). Several antisense constructs were tested; data from two constructs showing the most dramatic effects on NKLAM expression with little or no toxicity are presented. Electroporation was used to introduce ODN into cells, which facilitates entry and allows lower ODN concentrations to be used (44). By monitoring the expression of several RNA species over time, including IFN-γ and RPS3, we were able to determine that NKLAM antisense oligonucleotide treatment selectively decreased NKLAM RNA levels, while having no apparent effect on other gene expression. This reduction in NKLAM RNA expression was associated with a significant reduction in both NK and CTL cytolytic function. Further specificity of the antisense NKLAM treatment for cytotoxicity was shown by its lack of effect on NK cell growth or viability. Cumulatively, these results support the thesis that NKLAM is associated with NK and T cell lytic activity.

Subcellular fractions of NK3.3 cells were obtained by density gradient centrifugation. Immunoblotting of fractions using Abs to NKLAM and granzyme B indicate that these two proteins precisely comigrate and peak at a density characteristic of cytoplasmic granules. These data plus the putative transmembrane domains predicted for NKLAM suggest that NKLAM is present within granule membranes. Further fractionation studies and immunolocalization studies are in progress. We have shown that small amounts of NKLAM protein are detectable in unstimulated NK3.3 cells; IL-2 stimulation increases NKLAM protein levels 10- to 20-fold (Fig. 6). Our working hypothesis is that preformed NKLAM is responsible for basal NK activity, but higher levels must be produced for cytokine-mediated, enhanced NK cytolysis. A similar scenario might be envisaged for CTL function, where Ag stimulation would induce higher levels of NKLAM expression, and therefore increase cytotoxicity. The antisense experiments suggest that NKLAM protein is relatively short lived and can be significantly depleted within 18–24 h. This would explain the dramatic reduction in NK and CTL function after antisense treatment.

High level expression of NKLAM mRNA is limited to activated or actively killing NK cells, CTL, and macrophages. In freshly isolated monocytes, the initially high level of NKLAM mRNA (which may in part be due to the mechanical stimulation of these cells by adherence, followed by scraping) can be further increased by stimulation with IFN-β (Fig. 7 A) or with IFN-γ and LPS (data not shown). While NK cells, CTL, and macrophages have distinct lineages and immunological properties, they share the ability to kill. The major mechanism of NK- and CTL-mediated cytolysis is by the Ca2+-dependent release of perforin and granzymes from cytoplasmic granules. A second lytic pathway functional in NK cells and CTL is mediated by Fas-FasL interactions. It is highly likely that NKLAM is at least involved in the perforin-dependent mechanism of killing due to its location in the cytolytic granules and since K562, the tumor target used in these studies, does not express Fas. It remains to be determined whether NKLAM may also play a role in Fas-dependent lysis.

Monocyte/macrophage killing of tumor targets appears to be mediated by TNF-α and/or nitric oxide production (52, 53, 54, 55). TNF-mediated killing results from interaction between insoluble TNF on the surface of the macrophage and the TNF receptor on the target cell. It is of interest that the TNF receptor (types I and II) and Fas are members of the same family of molecules with three or four modules of cysteine-rich domains in their N-termini (56, 57). Although the cysteine arrangement in NKLAM is different from that of the TNF receptor family, the suggestion of a distant similarity in structure, and perhaps function, is nevertheless intriguing.

In summary, the restricted expression of NKLAM mRNA to killer cells (NK cells, CTL, and macrophages), the strong correlation of NKLAM RNA expression with cytolytic function and up-regulation by cytokines that enhance cytolysis, the localization of NKLAM protein to NK cytolytic granules, and the selective inhibition of cytokine-enhanced cytolysis by NKLAM antisense oligonucleotide-treated NK cells and CTL, strongly indicate an important role for NKLAM in cell-mediated cytotoxicity. Further functional characterization of NKLAM should increase our understanding of the killing mechanism used by NK and other cytolytic cells, which is critical for designing better means of immunotherapeutic intervention and perhaps, ultimately, the prevention of cancer and viral diseases.

We thank James Crouch, J. Catherine Cone, and Kathleen Hayes for their assistance and Dr. Richard Hoover for his constructive comments.

1

This work was supported by grants from the American Cancer Society (IM-618), the National Institutes of Health (CA62201), and the Veterans Administration.

3

Abbreviations used in this paper: NKLAM, NK lytic-associated molecule; RPS3, ribosomal protein S3; AbN, Ab to the N-terminus of NKLAM; AbC, Ab to the C-terminus of NKLAM; S, sense; AS, antisense; ODN, oligodeoxynucleotides.

1
Herberman, R. B., J. R. Ortaldo.
1981
. Natural killer cells: their role in defenses against disease.
Science
214
:
24
2
Trinchieri, G..
1989
. Biology of natural killer cells. F. Dixon, ed. In
Advances in Immunology
Vol. 47
:
187
Academic Press, Orlando.
3
Kuribayashi, K., S. Gillis, D. E. Kern, C. S. Henney.
1981
. Murine NK cell cultures: effects of interleukin-2 and interferon on cell growth and cytotoxic reactivity.
J. Immunol.
126
:
2321
4
Trinchieri, G., M. Matsumoto-Kobayashi, S. C. Clark, J. Seehra, L. London, B. Perussia.
1984
. Response of resting human peripheral blood natural killer cells to interleukin 2.
J. Exp. Med.
160
:
1147
5
Kobayashi, M., L. Fitz, M. Ryan, R. M. Hewick, S. C. Clark, S. Chan, R. London, F. Sherman, B. Perussia, G. Trinchieri.
1989
. Identification and purification of natural killer cell stimulatory factor (NKSF): a cytokine with multiple biologic effects on human lymphocytes.
J. Exp. Med.
170
:
827
6
Carson, W. E., J. G. Giri, M. J. Lindemann, M. L. Linett, M. Ahdieh, R. Paxton, D. Anderson, J. Eisenmann, K. Grabstein, M. A. Caligiuri.
1994
. Interleukin (IL) 15 is a novel cytokine that activates human natural killer cells via components of the IL-2 receptor.
J. Exp. Med.
180
:
1395
7
Phillips, J. H., L. L. Lanier.
1986
. Dissection of the lymphokine-activated killer phenomenon: relative contribution of peripheral blood natural killer cells and T lymphocytes to cytolysis.
J. Exp. Med.
164
:
814
8
Rosenberg, S. A., M. T. Lotze, L. M. Muul, A. E. Chang, F. P. Avis, A. Leitman, W. M. Linehan, C. N. Robertson, R. E. Lee, J. T. Rubin, et al
1987
. A progress report on the treatment of 157 patients with advanced cancer using lymphokine-activated killer cells and interleukin-2 or high-dose interleukin-2 alone.
N. Engl. J. Med.
316
:
889
9
Rosenberg, S. A., J. C. Yang, S. L. Topalian, D. J. Schwartzentruber, J. S. Weber, D. R. Parkinson, L. A. Seipp, J. H. Einhorn, D. E. White.
1994
. Treatment of 283 consecutive patients with metastatic melanoma or renal cell cancer using high dose bolus IL-2.
JAMA
271
:
907
10
Rosenstein, M., S. E. Ettinghausen, S. A. Rosenberg.
1986
. Extravasation of intravascular fluid mediated by the systemic administration of recombinant interleukin-2.
J. Immunol.
137
:
1735
11
Belldegrun, A., D. E. Webb, H. A. Austin, S. M. Steinberg, D. E. White, W. M. Linehan, S. A. Rosenberg.
1987
. Effects of interleukin-2 on renal function in patients receiving immunotherapy for advanced cancer.
Ann. Intern. Med.
106
:
817
12
Richards, J. M., M. T. Lotze.
1992
. IL-2 therapy: current status and future directions.
Contemp. Oncol.
92
:
34
13
Ortaldo, J. R., W. Phillips, K. Wasserman, R. B. Herberman.
1980
. Effects of metabolic inhibitors on spontaneous and interferon-boosted human natural killer cell activity.
J. Immunol.
125
:
1839
14
Kornbluth, J., L. Searfoss.
1990
. c-myb protooncogene levels correlate with interleukin-2 and interferon-β-induced augmentation of natural killer activity. R. E. Schmidt, ed.
Natural Killer Cells: Biology and Clinical Application
212
Karger, Basel.
15
Dahl, C. A., R. P. Schall, H. He, S. Cairns.
1992
. Identification of a novel gene in activated natural killer cells and T cells.
J. Immunol.
148
:
597
16
Salcedo, T. W., L. Azzoni, S. F. Wolf, B. Perussia.
1993
. Modulation of perforin and granzyme messenger RNA expression in human natural killer cells.
J. Immunol.
151
:
2511
17
Robertson, M. J., M. A. Caligiuri, T. J. Manley, H. Levine, J. Ritz.
1990
. Human natural killer adhesion molecules: differential expression after activation and participation in cytolysis.
J. Immunol.
145
:
3194
18
Kornbluth, J., N. Flomenberg, B. Dupont.
1982
. Cell surface phenotype of a cloned line of human natural killer cells.
J. Immunol.
129
:
2831
19
Leiden, J. M., K. M. Gottesdiener, J. Quertermous, L. Coury, R. A. Bray, L. Gottschalk, H. Gebel, J. G. Seidman, J. L. Strominger, A. L. Landay, et al
1988
. T-cell receptor gene rearrangement and expression in human natural killer cells: natural killer activity is not dependent on the rearrangement and expression of T-cell receptor α, β, or γ genes.
Immunogenetics
27
:
231
20
Kornbluth, J., R. G. Hoover.
1988
. Changes in gene expression associated with IFN-β and IL-2-induced augmentation of human natural killer cell function.
J. Immunol.
141
:
3234
21
Kornbluth, J., D. B. Wilson. Monoclonal antibodies directed against HLA molecules affect the lytic and proliferative behavior of a cloned line of human natural killer cells.
Hum. Immunol.
11
:
1985a
239
22
Kornbluth, J., B. Spear, S. S. Raab, D. B. Wilson. Evidence for the role of class I and class II HLA antigens in the lytic activity of a cloned line of human natural killer cells.
J. Immunol.
134
:
1985b
728
23
Kanof, M. E..
1991
. Isolation of T cells using rosetting procedures. J. E. Coligan, and A. M. Kruisbeek, and D. H. Margulies, and E. M. Shevach, and W. Strober, eds.
Current Protocols in Immunology
7.2.1.
Greene Publishing Associates, New York.
24
Chomczynski, P., N. Sacchi.
1987
. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction.
Anal. Biochem.
162
:
156
25
Devereux, J., P. Haeberli, O. Smithies.
1984
. A comprehensive set of sequence analysis programs for the VAX.
Nucleic Acids Res.
12
:
387
26
Altshul, S. F., W. Gish, W. Miller, E. W. Myers, D. J. Lipman.
1990
. Basic local alignment search tool.
J. Mol. Biol.
215
:
403
27
Pearson, W. R., D. J. Lipman.
1988
. Improved tools for biological sequence comparison.
Proc. Natl. Acad. Sci. USA
85
:
2444
28
Laemmli, U. K..
1970
. Cleavage of structural proteins during the assembly of the head of bacteriophage T4.
Nature
227
:
680
29
Grigoriev, V. G., R. Thweatt, E. J. Moerman, S. Goldstein.
1996
. Expression of senescence-induced protein WS3–10 in vivo and vitro.
Exp. Gerontol.
31
:
145
30
Towbin, H., T. Staehelin, J. Gordon.
1979
. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications.
Proc. Natl. Acad. Sci USA
76
:
4350
31
Shi, L., R. P. Kraut, R. Aebersold, A. H. Greenberg.
1992
. A natural killer cell granule protein that induces DNA fragmentation and apoptosis.
J. Exp. Med.
175
:
553
32
Kozak, M..
1984
. Compilation and analysis of sequences upstream from the translational start site in eukaryotic mRNAs.
Nucleic. Acids Res.
12
:
857
33
Kozak, M.. An analysis of vertebrate mRNA sequences: intimations of translational control.
J. Cell. Biol.
115
:
1991a
887
34
Lopez-Alanon, D. M., J. del Mazo.
1995
. Cloning and characterization of genes expressed during gametogenesis of female and male mice.
J. Reprod. Fertil.
103
:
323
35
von Heijne, G..
1986
. A new method for predicting signal sequence cleavage sites.
Nucleic Acids Res.
14
:
4683
36
Rao, J. K. M., P. Argos.
1986
. A conformational preference parameter to predict helices in integral membrane proteins.
Biochim. Biophys. Acta
869
:
197
37
Hofmann, K., W. Stoffel.
1993
. TMbase-A database of membrane spanning proteins segments.
Biol. Chem. Hoppe-Seyler
347
:
166
38
Bairoch, A..
1991
. PROSITE: a dictionary of sites and patterns in proteins.
Nucleic Acids Res.
19
:
2241
39
Freemont, P. S..
1993
. The RING finger: a novel protein sequence motif related to the zinc finger.
Ann. NY Acad. Sci.
684
:
174
40
Berg, J. M., Y. Shi.
1996
. The galvanization of biology: a growing appreciation for the roles of zinc.
Science
271
:
1081
41
Saurin, A. J., K. L. B. Borden, M. N. Boddy, P. Freemont.
1996
. Does this have a familiar RING?.
Trends Biochem. Sci.
21
:
208
42
Pogue-Geile, K., J. R. Geisner, M. Shu, C. Miller, I. G. Wool, A. I. Meisler, J. M. Pipas.
1991
. Ribosomal protein genes are overexpressed in colorectal cancer: isolation of a cDNA clone encoding the human S3 ribosomal protein.
Mol. Cell. Biol.
11
:
3842
43
Ye, J., J. R. Ortaldo, K. Conlon, R. Winkler-Pickett, H. A. Young.
1995
. Cellular and molecular mechanisms of IFN-γ production induced by IL-2 and IL-12 in a human NK cell line.
J. Leukocyte Biol.
58
:
225
44
Bergan, R., Y. Connell, B. Fahmy, L. Neckers.
1993
. Electroporation enhances c-myc antisense oligodeoxynucleotide efficacy.
Nucleic Acids Res.
21
:
3567
45
Young, J. D.-E., L. G. Leong, C.-C. Liu, A. Damiano, D. A. Wall, Z. A. Cohn.
1986
. Isolation and characterization of a serine esterase from cytolytic T cell granules.
Cell
47
:
183
46
Kozak, M.. Structural features in eukaryotic mRNAs that modulate the initiation of translation.
J. Biol. Chem.
266
:
1991b
19867
47
Talento, A., M. Nguyen, S. Law, J. K. Wu, M. Poe, J. T. Blake, M. Patel, T.-J. Wu, C. L. Manyak, M. Silberklang, et al
1992
. Transfection of mouse cytotoxic T lymphocytes with an antisense granzyme A vector reduces lytic activity.
J. Immunol.
149
:
4009
48
Bochan, M. R., W. S. Goebel, Z. Brahmi.
1995
. Stably transfected antisense granzyme B and perforin constructs inhibit human granule-mediated lytic ability.
Cell. Immunol.
164
:
234
49
Giardina, S. L., K. S. Anderson, J. T. Sayers, H. W. Chambers, A. G. Palumbo, H. A. Young, J. R. Ortaldo.
1995
. Selective loss of NK cytotoxicity in antisense NK-TR1 rat LGL cell lines: abrogation of antibody-independent tumor and virus-infected target cell killing.
J. Immunol.
154
:
80
50
Ortaldo, J. R., A. T. Mason, L. H. Mason, R. T. Winkler-Pickett, P. Gosselin, S. K. Anderson.
1997
. Selective inhibition of human and mouse natural killer tumor recognition using retroviral antisense in primary natural killer cells. Involvement with MHC class I killer cell inhibitory receptors.
J. Immunol.
158
:
1262
51
Stein, C. A..
1995
. Does antisense exist? It may, but only under very special circumstances.
Nat. Med.
1
:
1119
52
Cox, G. W., G. Melillo, U. Chattopadhyay, D. Mullet, R. H. Fertel, L. Varesio.
1992
. Tumor necrosis factor-α dependent production of reactive nitrogen intermediates mediate IFN-γ plus IL-2 induced murine macrophage tumoricidal activity.
J. Immunol.
149
:
3290
53
Wynn, T. A., Y. R. Freund, D. M. Paulnock.
1992
. TNFα differentially regulates Ia antigen expression and macrophage tumoricidal activity in two murine macrophage cell lines.
Cell. Immunol.
140
:
184
54
Lorsbach, R. B., W. J. Murphy, C. J. Lowenstein, S. H. Snyder, S. W. Russell.
1993
. Expression of the nitric oxide synthase gene in mouse macrophages activated for tumor cell killing: molecular basis for the synergy between interferon-γ and lipopolysaccharide.
J. Biol. Chem.
268
:
1908
55
Cui, S., J. S. Reichner, R. B. Mateo, J. E. Albina.
1994
. Activated murine macrophages induce apoptosis in tumor cells through nitric oxide-dependent or -independent mechanisms.
Cancer Res.
54
:
2462
56
Gruss, H. J., S. K. Dower.
1995
. Tumor necrosis factor ligand superfamily: Involvement in the pathology of malignant lymphomas.
Blood
85
:
3378
57
Baker, S. J., E. P. Reddy.
1996
. Transducers of life and death: TNF receptor superfamily and associated proteins.
Oncogene
12
:
1