| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Pathology, St. Louis University School of Medicine, and Veterans Administration Medical Center, St. Louis, MO 63106
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, ß, and 
), 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-ß 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-ß. We have previously found that the protooncogene c-myb is up-regulated by both IL-2 and IFN-ß 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-ß-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-ß-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
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
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 x 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-ß was a gift from Triton Biosciences (Alameda, CA).
Isolation and purification of peripheral blood monocytes, T cells, and NK cells
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-ß (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 x 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 x 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+.
Construction of cDNA libraries
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-ß. 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 
ZAP 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-ß-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.
Sequencing
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.
Cytolytic assays
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)] x 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%.
Northern blots
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 x 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- 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 protein expression
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.
NKLAM Ab production and purification
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 and Western blotting
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 x g in a refrigerated microcentrifuge. The supernatant was designated the cytoplasmic extract and used for subsequent studies. Extracts equivalent to 20 x 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).
NKLAM antisense experiments
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 x 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 x
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 x 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 x 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).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The strategy for cloning cytolysis-specific genes consisted of selecting genes expressed at elevated levels in IFN-ß-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-ß, 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-ß, 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-ß.
A random-primed library in Zap II was constructed using cDNA
derived from NK3.3 cells stimulated with 5000 U/ml of IFN-ß for
4 h. Plaques (1 x 106) were
differentially screened using radiolabeled cDNA derived from poly(A)
mRNA from starved and IFN-ß-treated NK3.3 cells. Clones were selected
that hybridized strongly to RNA from IFN-ß-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.
Characterization of NKLAM
Clones with unique sequences were used as probes to screen RNA
from IL-2, IFN-ß, 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-ß-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).
|
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.
|
|
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.
Regulation of NKLAM expression and cytolysis in NK3.3
The kinetics of induction of NKLAM mRNA expression and
corresponding cytolytic function of NK3.3 cells after stimulation with
IL-2 and IFN-ß 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-ß (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. 4C).
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. 5C). 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.
|
|
; it was used to monitor NK cell activation. Ye et al.
(43) reported that NK3.3 cells produce IFN- after IL-2
treatment. Here, we show that both IL-2 and IFN-ß induce the
expression of IFN- mRNA in NK3.3 cells. (Figs. 4 and 5) The kinetics
of induction of IFN- mRNA in NK3.3 cells upon IL-2 and IFN-ß
stimulation closely parallel the kinetics of induction of NKLAM. This
suggests that there may be a common regulatory pathway for IFN- 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, lanes 1 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.
|
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. 7A. 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-ß-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-ß 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-ß-stimulated NK 3.3 cells; bar
3). These levels are increased 3- to 5-fold further by 24-h
IFN-ß 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.
|
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-ß. 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). Potential role of NKLAM in cytolysis
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- 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- mRNA expression.
Fig. 9C 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. 10A). NKLAM
AS ODN treatment resulted in a 60% decrease in CTL activity. A
corresponding, selective decrease in NKLAM RNA levels was seen (Fig. 10B). 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.
NKLAM colocalizes with granzyme B to cytoplasmic granules
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 ß-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.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
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-ß 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- mRNA in NK3.3 cells, resulting in a
significant accumulation and corresponding increase in IFN- 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-ß-treated cells (data not shown). Therefore, the increased
steady state levels of NKLAM mRNA in response to IL-2 and IFN-ß 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- 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-ß (Fig. 7A) or with IFN- 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- 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.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jacki Kornbluth, Department of Pathology, St. Louis University School of Medicine, 1402 South Grand Boulevard, St. Louis, MO 63104. E-mail address:
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.
Received for publication April 8, 1999. Accepted for publication May 28, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, ß, or genes. Immunogenetics 27:231.[Medline]
production induced by IL-2 and IL-12 in a human NK cell line. J. Leukocyte Biol. 58:225.[Abstract]
dependent production of reactive nitrogen intermediates mediate IFN- plus IL-2 induced murine macrophage tumoricidal activity. J. Immunol. 149:3290.[Abstract]
differentially regulates Ia antigen expression and macrophage tumoricidal activity in two murine macrophage cell lines. Cell. Immunol. 140:184.[Medline]
and lipopolysaccharide. J. Biol. Chem. 268:1908.This article has been cited by other articles:
|
|
 |
|
  J. M. Fortier and J. Kornbluth NK Lytic-Associated Molecule, Involved in NK Cytotoxic Function, Is an E3 Ligase. J. Immunol., June 1, 2006; 176(11): 6454 - 6463. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Wex, F. Buhling, H. Wex, D. Gunther, P. Malfertheiner, E. Weber, and D. Bromme Human Cathepsin W, a Cysteine Protease Predominantly Expressed in NK Cells, Is Mainly Localized in the Endoplasmic Reticulum J. Immunol., August 15, 2001; 167(4): 2172 - 2178. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Shallal and J. Kornbluth CD9 expression enhances the susceptibility of myeloma cell lines to cell-mediated cytolysis Blood, July 1, 2000; 96(1): 224 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Oka, K. Udaka, A. Tsuboi, O. A. Elisseeva, H. Ogawa, K. Aozasa, T. Kishimoto, and H. Sugiyama Cancer Immunotherapy Targeting Wilms' Tumor Gene WT1 Product J. Immunol., February 15, 2000; 164(4): 1873 - 1880. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
