Impaired NK Cytolytic Activity and Enhanced

Tumor Growth in NK Lytic-Associated

This information is current as
of January 26, 2011
Molecule-Deficient Mice
Richard G. Hoover, Gail Gullickson and Jacki Kornbluth
J Immunol 2009;183;6913-6921; Prepublished online 13
November 2009;
doi:10.4049/jimmunol.0901679
http://www.jimmunol.org/content/183/11/6913

Downloaded from www.jimmunol.org on January 26, 2011

References This article cites 37 articles, 17 of which can be accessed free at:
http://www.jimmunol.org/content/183/11/6913.full.html#ref-list-1
Subscriptions Information about subscribing to The Journal of Immunology is online at
http://www.jimmunol.org/subscriptions
Permissions Submit copyright permission requests at
http://www.aai.org/ji/copyright.html
Email Alerts Receive free email-alerts when new articles cite this article. Sign up at
http://www.jimmunol.org/etoc/subscriptions.shtml/

The Journal of Immunology is published twice each month by

The American Association of Immunologists, Inc.,
9650 Rockville Pike, Bethesda, MD 20814-3994.
Copyright ©2009 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
 The Journal of Immunology

Impaired NK Cytolytic Activity and Enhanced Tumor Growth

in NK Lytic-Associated Molecule-Deficient Mice1,2

Richard G. Hoover,† Gail Gullickson,† and Jacki Kornbluth3*†

NK lytic-associated molecule (NKLAM) is a protein involved in the cytolytic function of NK cells. It is weakly expressed in resting
NK cells but upon target cell stimulation or after incubation with cytokines that enhance NK killing, NKLAM mRNA levels
increase and protein is synthesized and is targeted to cytoplasmic granule membranes. We have previously shown that NKLAM
plays a role in perforin/granzyme-mediated cytolysis in vitro. To further investigate the function of NKLAM in NK cell-mediated
cytotoxicity, we generated, by gene targeting, NKLAM-deficient mice. These mice have normal numbers of NK cells and other
lymphoid populations in the spleen. They also have no alterations in NK maturation or NK receptor repertoire. NK cells from
NKLAM-deficient and WT mice have comparable amounts of perforin, granzyme B, and lysosomal membrane-associated protein
1 (CD107a) in their cytotoxic granules and comparable levels of granule exocytosis are induced by PMA and calcium ionophore
A23187. However, NKLAM-deficient NK cells display significantly less NK cytotoxic activity in vitro than WT NK cells. They also
secrete less IFN-␥ upon target cell stimulation, In addition, NKLAM-deficient mice exhibit greater numbers of pulmonary me-

Downloaded from www.jimmunol.org on January 26, 2011

tastases after i.v. injection with B16 melanoma cells. These studies indicate that NKLAM-deficient mice have diminished capacity
to control tumor metastases and support the role for NKLAM in NK function both in vitro and in vivo. The Journal of Immu-
nology, 2009, 183: 6913– 6921.

N atural killer cells are lymphocytes of the innate immune
system that play an important role in early host defense
against tumors and infectious agents (1). NK cells can
induce target cell death by a variety of mechanisms. However, the
primary mechanism of killing is the release of cytotoxic granules
containing perforin and granzymes (2– 4). It has long been known
that NK-mediated killing can be augmented by cytokines, includ-
ing IL-2, IL-12, IL-15, and IFN (5– 8). In studies to characterize
additional proteins associated with the cytolytic process, we iden-
tified a novel protein whose expression was highly increased upon
cytokine stimulation of NK cells (9). This protein was named NK
lytic-associated molecule (NKLAM).4
NKLAM is found in the cytolytic granule membranes of NK
cells and CTL. It is weakly expressed in resting NK cells and
unlike perforin and granzymes, there is little to no preformed
NKLAM mRNA or protein in NK cells. Upon target cell stimu-
lation or after incubation with cytokines that enhance NK killing,
NKLAM mRNA levels in NK cells increase and protein is syn-
*Department of Veterans Affairs Medical Center, St. Louis, MO 63106; and †De-
partment of Pathology, Saint Louis University School of Medicine, St. Louis, MO
63104
Received for publication May 27, 2009. Accepted for publication September
21, 2009.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This material is based upon work supported by a merit review grant from the De-
partment of Veterans Affairs, Veterans Health Administration, Office of Research and
Development, Biomedical Laboratory Research and Development.
2
The views expressed in this article are those of the authors and do not necessarily
reflect the position or policy of the Department of Veterans Affairs or the U.S.
government.
3 membrane-associated protein 1 (LAMP-1; CD107a) in their cyto-
Address correspondence and reprint requests to Dr. Jacki Kornbluth, Department of
Pathology, St. Louis University School of Medicine, DRC 323, 1100 South Grand
Boulevard, St. Louis, MO 63104. E-mail address: kornblut@slu.edu
4
Abbreviations used in this paper: NKLAM, NK lytic-associated molecule; IBR, in
between RING; KO, knockout; LAMP-1, lysosomal membrane-associated protein 1;
neo, neomycin resistance; poly(I:C), polyinosinic:polycytidylic acid; RING, really
interesting new gene; WT, wild type; TK, thymidine kinase.
thesized and becomes targeted to cytoplasmic granule membranes.
Studies have shown a role for NKLAM in perforin/granzyme-me-
diated cytolysis. Treatment of NK cells or CTL with NKLAM
antisense oligonucleotides inhibits their killing activity, indicating
that NKLAM participates in the cytotoxic process (9, 10).
To establish what function NKLAM plays in NK-mediated kill-
ing, we focused our attention on the physical characteristics of
NKLAM protein. NKLAM contains three cysteine-rich domains
that form a really interesting new gene (RING)-in between RING
(IBR)-RING domain structure. This tripartite domain is also col-
lectively known as a TRIAD or RBR (11, 12). There are currently
over 140 RING-IBR-RING containing proteins in a wide variety
of organisms, from yeast to human (13). Many of these proteins
are associated with diverse protein-protein interactions. A subset
of these proteins has been found to possess E3 ubiquitin ligase
activity (12, 13). E3 ubiquitin ligases regulate the ubiquitination of
proteins, catalyzing the transfer of activated ubiquitin from an E2
ubiquitin-conjugating enzyme to its substrate (14, 15). Multiple E3
ubiquitin ligases have been shown to have important roles in a
variety of immune functions (16). For example, GRAIL is asso-
ciated with the induction of anergy in CD4⫹ T cells (17). Triad3A
targets TLRs for ubiquitination and subsequent degradation (18).
We have identified NKLAM as an E3 ubiquitin ligase but have not
yet determined how this enzymatic activity results in enhanced NK
cytolytic function (19).
To further investigate the role of NKLAM in cell-mediated cy-
totoxicity, NKLAM-deficient mice were generated. These mice
appear normal and are fertile. They have normal numbers of NK
cells in the spleen. NK cells from NKLAM knockout (KO) mice
have comparable amounts of perforin, granzyme B, and lysosomal
plasmic granules as NK cells from wild-type (WT) mice. How-
ever, these NKLAM-deficient NK cells display significantly less
NK activity in vitro than WT cells and produce less IFN-␥ upon
target cell stimulation. In addition, NKLAM-deficient mice exhibit
enhanced numbers of pulmonary metastases after injection with

www.jimmunol.org/cgi/doi/10.4049/jimmunol.0901679
6914 ENHANCED TUMOR METASTASIS IN NKLAM-DEFICIENT MICE
B16 melanoma cells. These studies indicate a significant role for
NKLAM in NK cytotoxic function in vitro and in vivo.
Materials and Methods
Constructs
NKLAM genomic clones were isolated from a 129 SvJ mouse liver
genomic library (␭ Fix II; Stratagene) by hybridization using mouse
NKLAM cDNA clones as previously described (10). Five overlapping
phage clones were obtained and the exon/intron boundaries were charac-
terized by PCR and sequencing. The targeting construct was generated
using the pKO scrambler NTKV-1905 vector (Stratagene). This vector
contains a neomycin resistance (neo) cassette for positive selection and a
HSV thymidine kinase (TK) cassette outside the region of homology to
allow for negative selection. The 5⬘ arm of the targeting construct was a
3.5-kb EcoRI/KpnI fragment located within intron 1 of NKLAM. The 3⬘
arm was a 6- kb EcoRI/EcoRI fragment extending from intron 5 to intron
8. This produced a targeting vector of 15 kb, which upon homologous
recombination would replace exons 2–5 of NKLAM with the neo cassette.
This construct was linearized with NotI and electroporated into 129Sv/Ev
embryonic stem cells (inGenious Targeting Lab). The neo-expressing em-
bryonic stem cells were selected in medium containing G418; DNA from
G418-resistant clones was isolated and evaluated for homologous recom-
bination by a combination of Southern blotting and PCR. PCR with
forward primer SMINTR1D (inside intron 1 within the 5⬘ arm) and
reverse primer A1570 (inside exon 6 within the 3⬘ arm) generated a WT
product of 7.4 kb and a KO product of 3 kb (PCR 1). Primers were
SMINTR1D: GCTATTGTTTCAAGCTGTGAG and AM1570: CACCTC
CACGGCAAAGAGAAATGG. Clones showing the 3-kb KO band were
confirmed as recombinant by two additional PCR using neo primers com-
bined with NKLAM-specific primers on either side of the targeting arms
(data not shown).
Production of NKLAM-deficient mice
Two recombinant embryonic stem cell clones were microinjected into
C57BL/6 blastocysts by inGenious. Coat color chimeric males (90 –100%
agouti) were then mated with C57BL/6 female mice and the progeny were
screened for germline transmission of the KO allele. Crossing of heterozy-
gous mice produced viable homozygous KO offspring. Pups were geno-
typed using DNA extracted from tail tissue by PCR 1. Heterozygous mice
have subsequently been backcrossed for 11 generations onto a C57BL/6
background. C57BL/6 mice were obtained from the National Cancer In-
stitute. Early in the backcross process, progeny were selected for the
C57BL/6 NK gene complex alleles by PCR so that the NK cells from these
mice were genotypically C57BL/6 (20). Primers for NKRP1a were: for-
ward, TGCTTGACTCAAATTCATGAGG; reverse, ATTTTACTCGA
CACCAGGTTGA, B6 allele 110 nt, 129 allele 133 nt. Primers for Ly49a
were: forward, GCCTCTGAGGTTCATGTTTGATT; reverse, TTCCTTC
CCTGTTGAGACTG, B6 allele 340 nt, 129 allele 321 nt. We observed no
significant differences in the results obtained from non-backcrossed and
fully backcrossed NKLAM homozygous KO mice. All mice were main-
tained in specific pathogen-free barrier housing and all experiments were
conducted in accordance with institutional animal care and use guidelines.
Cell lines
YAC-1 and B16-F10 melanoma cells were purchased from the American
Type Culture Collection. The NK- sensitive mouse lymphoma YAC-1 cell
line was maintained in medium containing RPMI 1640 supplemented with
7.5% FBS and 1% glutamine. They were split twice a week and maintained
at a density of 2 ⫻ 105 cells/ml. B16 melanoma cells were propagated in
DMEM containing 7.5% FBS and 1% glutamine. Subconfluent adherent
cells were trypsinized and split twice a week.
Isolation of spleen cells
Spleens were harvested, rinsed in RPMI 1640, and gently homogenized
using the blunt end of a 3-ml syringe. Released spleen cells were pelleted
and then layered over a Lympholyte-M density gradient (Cedarlane Lab-
oratories/ Accurate Chemical Corporation) according to the manufacturers’
instructions to remove RBC, dead cells, and debris. Lymphocytes at the
interface were collected, washed twice with RPMI 1640, and counted.
These cells were analyzed by flow cytometry and/or used for the purifica-
tion of NK cells.
Flow cytometry
Flow cytometry was performed to analyze spleen cell subpopulations and
NK cells from WT and NKLAM-deficient mice. Briefly, tubes containing
2 ⫻ 105–2 ⫻ 106 cells in 0.05 ml of flow buffer (ice-cold PBS plus 1% FBS
plus 0.1% sodium azide) were incubated with 1 ␮l of Fc block (BD Pharm-
ingen) for 10 min at 4oC. Fluorescent-conjugated cell surface-specific pri-
mary Abs were then added and incubated for an additional 30 min at 4oC.
Cells were washed twice with flow buffer and then either fixed with 0.2 ml
of PBS plus 1% formaldehyde or prepared for intracellular staining using
the BD Cytofix/Cytoperm Plus Fixation/Permeabilization Kit with Golgi-
Stop according to the manufacturer’s directions. Cells were analyzed on a
BD Biosciences FACSCalibur flow cytometer. Abs purchased from BD
Biosciences were FITC-CD3 (no. 553061), Alexa Fluor 700-CD3 (no.
557984), allophycocyanin-NK1.1 (no. 550627), PE-Cy7-NK1.1 (no.
552878), PE-streptavidin (no. 13025D), PE-CD27 (no. 558754), PerCP-
CD45 (no. 557235), PE-CD49b (DX5; no. 553858), allophycocyanin-
CD62L (L-selectin; no. 553152), PE-CD8a (no. 553032), allophycocyanin-
CD4 (no. 553051), allophycocyanin-CD19 (no. 550992), FITC-Gr-1
(Ly-6G and Ly-6C; no. 553126), PE-CD11b (no. 557397), PE-Cy7-CD11b
(no. 553311), PE-CD25 (no. 553866), biotin-KLRG1 (no. 5500863), PE-
NKG2D (no. 558403), FITC-Ly49C/I (no. 553278), PE-Ly49A (no.
557424), and control Abs were FITC-IgG1 (no. 553971), PE-IgG2b (no.
553987), and allophycocyanin-IgG2a (no. 553932). Abs purchased from
eBioscience were FITC-NKp46 (no. 11-3351-80), PE-Ly49I (no.
12-5895), PE-granzyme B (no. 12-8899-41), and FITC-perforin (no. 11-
9392-82). Ly49H, Ly49A, and Ly49I Abs were generously provided by Dr.
M. Buller (St. Louis University). Samples were analyzed using FlowJo
software (Tree Star).
Downloaded from www.jimmunol.org on January 26, 2011
Purification of NK cells and stimulation
NK cells were isolated from spleen cells using magnetic CD49b (DX5)
microbeads, a MiniMACS Separator, and MS columns according to the
manufacturer’s recommendations (Miltenyi Biotec) (21). These cells were
routinely ⬎92% DX5 by FACS analysis. For cytokine stimulation, NK
cells were cultured at 5 ⫻ 106 cells/ml in RPMI 1640 containing 10% FBS,
1% glutamine, 5 ⫻ 10⫺5 M 2-ME, and 500 U/ml rIL-2. IL-2 was a gift
from Chiron. For some experiments, unseparated spleen cells were simi-
larly cultured in IL-2 in parallel.
NK cells were activated in vivo by i.p. injection of mice with 200 ␮g of
polyinosinic:polycytidylic acid (poly(I:C), sodium salt; Sigma-Aldrich).
Control mice were injected with PBS. After 24 h, spleen cells or purified
NK cells were assayed for NK activity in vitro.
Killing assays and granule exocytosis
Unstimulated and IL-2-stimulated purified NK cells were used as effectors
in cytotoxicity assays against the 51Cr-labeled NK-sensitive mouse lym-
phoma cell line YAC-1, as previously described (10). Assay conditions
were set up in triplicate wells at various E:T ratios in 96-well round-bottom
microtiter plates. After incubation for 4 h at 37oC, supernatants (0.1 ml)
were harvested and the amount of 51Cr release from YAC-1 target cells was
detected using a Packard Cobra II gamma counter. The percent specific
lysis of targets was calculated as previously described. The SD of triplicate
values never exceeded 10% in all experiments.
Granule exocytosis was monitored by the release of ␤-glucuronidase
using the fluorogenic substrate 4-methylumbelliferyl-␤-D-glucuronide (22).
Reagents used in this assay were purchased from Sigma-Aldrich. Four ⫻
105 spleen cells or 2.5 ⫻ 105 purified NK cells were incubated with 0.025
ml of RPMI 1640 containing 4% FBS in 96-well plates. PMA (40 ng/ml)
plus calcium ionophore A23187 (2 ␮g/ml) was used to induce granule
exocytosis. Total release was obtained by incubation of cells with 1%
Triton X-100. After 4 h at 37oC, cells were pelleted and supernatants were
harvested. In brief, 0.005 ml of supernatant was added to triplicate wells of
96-well black microtiter plates and mixed with 0.05 ml of ␤-glucuronidase
substrate (1.5 mM 4-methylumbelliferyl-␤-D-glucuronide in 0.05 M so-
dium acetate, 0.01% BSA, and 0.1% Triton X-100, pH 4). Plates were
incubated for 10 min at 37oC. The reaction was stopped by the addition of
0.175 ml of 0.25 M sodium tetraborate. The fluorescence intensity was
measured with excitation at 355 nm and emission of 460 nm in a Wallac
Victor 1420 multilabel counter.
IFN-␥ production
Unstimulated and IL-2-stimulated purified NK cells were cocultured with
YAC-1 target cells in triplicate wells at E:T ratios of 25:1 and 10:1, re-
spectively, in 96-well round-bottom microtiter plates. After incubation for
4 h at 37oC, supernatants (0.1 ml) were harvested and the amount of IFN-␥
was quantitated by ELISA (Endogen and Pierce Biotechnology) according
to the manufacturers’ protocol.
The Journal of Immunology 6915

FIGURE 1. Generation and characterization of NKLAM-deficient mice. A, The mouse NKLAM targeting construct replaced exons 2–5 of the NKLAM
gene, encoding the second and third cysteine-rich domains and two of the three predicted transmembrane domains, with the neomycin resistance (neo) gene.
The targeting vector also contained a TK cassette outside the region of homology. This targeting construct was transfected into 129Sv/Ev embryonic stem
cells and neo-expressing clones were selected in medium containing G418. Homologous recombinant clones were identified by a combination of Southern

Downloaded from www.jimmunol.org on January 26, 2011

blotting and PCR. B, Progeny were genotyped from tail DNA using multiple PCR. In the representative PCR shown on the left, the primers are on either
side of knocked out exons 2–5 of the NKLAM gene. The WT band is 7.4 kb while the KO band is 3 kb. On the right, RT-PCR confirmed the absence of
NKLAM mRNA in IL-2-stimulated spleen cells in NKLAM⫺/⫺ mice. These PCR primers were located within the KO region of NKLAM. ␤-Actin served
as a positive PCR control.

Immunoblot analysis
Rat monoclonal anti-LAMP-1 (CD107a) mAb 1D4B, provided by Dr. J. T.
August, was obtained from the Developmental Studies Hybridoma Bank
developed under the auspices of the National Institute of Child Health and
Human Development and maintained by the University of Iowa (Depart-
ment of Biological Sciences, Iowa City, IA) (23). It was used at 1/100.
Anti-perforin Ab (no. 3693, 1/1,000) and anti-granzyme B Ab (no. 4275,
1/1000) were purchased from Cell Signaling. Anti-␤-actin mAb clone
AC-15 (1/10,000) was purchased from Sigma-Aldrich. NKLAM mAb 35
was produced in our laboratory and used at 35 ␮g. Immunoblots were
performed as previously described (19). Briefly, cell extracts were prepared
from purified NK cells from NKLAM-deficient and WT mice by lysing
cells in buffer containing 20 ␮M PIPES, 4% sucrose, 100 mM NaCl, and
0.5% Nonidet P-40 plus protease inhibitors (25 ␮g/ml aprotinin, leupeptin,
pepstatin, 2 mM benzamidine, 1 mM PMSF). Lysates were sonicated twice
for 10 s each, then mixed with 2⫻ SDS loading buffer. Samples were
boiled for 20 min and loaded onto 7.5% SDS-PAGE at 100 –150 V. Pro-
teins were transferred to polyvinylidene difluoride membranes, incubated
with blocking buffer for various times (10 mM Tris, 150 mM NaCl, 0.2%
Tween 20, and 5% blotto) and then incubated with primary Ab. Secondary
Abs were HRP-conjugated goat anti-mouse IgG (1/10,000; Bio-Rad), goat
anti-rabbit (1/2,000; Cell Signaling), or donkey anti-rat IgG (1/5,000; Mil-
lipore). Membranes were then placed in Supersignal Chemiluminescence
solution (Pierce) and exposed to film.
Lung metastasis studies
B16 melanoma cells were injected i.v. through the tail vein into NKLAM-
deficient KO and WT mice. Cells were washed with PBS and each mouse
was injected with 300,000 cells in 0.2 ml of PBS. Fifteen days later, ani-
mals were sacrificed and lungs were harvested and fixed in Fekete’s solu-
tion. Individual lobes were dissected and black B16 melanoma metastatic
nodules were enumerated on each lobe surface in a blinded fashion. Data
were analyzed using a two-tailed Student’s t test.
Results
Generation of NKLAM-deficient mice
Genomic clones containing mouse NKLAM were isolated from a
129SvJ library. The NKLAM gene contains nine exons, spanning
over 29 kb. It was not possible to knock out exon 1, containing the
start codon, because intron 1 is extremely large, over 14 kb. There-
fore, a targeting vector was constructed in which exons 2–5 (⬃6
kb) were replaced with a neo cassette (Fig. 1A). This removed the
second and third cysteine-rich domains of NKLAM, which to-
gether form a RING-IBR-RING structure, important in the E3
ubiquitin ligase activity of NKLAM. It also removed two of the
three predicted transmembrane domains. This construct was trans-
fected into 129Sv/Ev embryonic stem cells. The neo clones were
selected in medium containing G418; homologous recombinant
clones were identified by a combination of Southern blotting and
PCR. Two recombinant clones were subsequently microinjected
into C57BL/6 blastocysts. Chimeric heterozygotes that displayed
germline transmission were intercrossed to produce NKLAM⫺/⫺
progeny (Fig. 1B). Heterozygotes were simultaneously back-
crossed to C57BL/6 mice for 11 generations to generate NKLAM-
deficient mice on a C57BL/6 background. Early in the backcross
process, heterozygous progeny were selected for the C57BL/6 NK
gene complex alleles by PCR so that the NK cells from these mice
were genotypically C57BL/6. Progeny were evaluated by a com-
bination of both Southern blotting and PCR. RT-PCR confirmed
the absence of NKLAM mRNA in IL-2-stimulated spleen cells
from NKLAM⫺/⫺ mice compared with heterozygous (⫹/⫺) and
WT (⫹/⫹) mice (Fig. 1B).
Homozygous NKLAM-deficient (⫺/⫺) mice were born at the
expected Mendelian ratio. Overall, the mice appeared normal and
were fertile. Histological examination of the major organs of 8-wk-
old NKLAM⫺/⫺ and heterozygous (⫹/⫺) mice revealed no overt
abnormalities by light microscopy. NKLAM⫺/⫺ mice were ob-
served for over 18 mo. They developed normally and there was no
obvious difference in gross morphology, size, or body weight be-
tween WT and NKLAM-deficient mice during the 18-mo period
(data not shown).
Normal distribution of lymphoid cells in the spleens of
NKLAM-deficient mice
The distribution of lymphoid cells in the spleens of NKLAM⫺/⫺
and NKLAM⫹/⫹ mice was analyzed by flow cytometry. There was
no significant difference in the absolute numbers of lymphoid cells
in NKLAM-deficient and WT mice; the total number of spleen
cells was 87.5 ⫾ 9.2 ⫻ 106 in NKLAM-deficient mice and
6916 ENHANCED TUMOR METASTASIS IN NKLAM-DEFICIENT MICE

Table I. Analysis of spleen cell subpopulations in NKLAM-deficient with NKLAM antisense oligonucleotides reduces both NKLAM
micea levels and NK-mediated cytotoxicity, indicating a role for
NKLAM in NK killing. Therefore, NK cells from NKLAM-defi-
Cell Type Designation NKLAM⫹/⫹ NKLAM⫺/⫺ cient mice were predicted to have lower NK activity. To test this
T cell CD3⫹ 36.9 ⫾ 3.1 33.6 ⫾ 2.6 hypothesis, DX5⫹ NK cells from the spleens of NKLAM⫺/⫺,
CD4⫹ 19.6 ⫾ 1.7 19.9 ⫾ 1.5 ⫹/⫺, and ⫹/⫹ WT mice were isolated by magnetic bead column
CD8⫹ 14.8 ⫾ 2.0 11.8 ⫾ 2.5 purification. Total numbers of NK cells were comparable among
Activated T CD3⫹CD25⫹ 3.2 ⫾ 0.9 3.1 ⫾ 0.8 the three groups. Freshly isolated NK cells were tested for killing
B cell CD19⫹ 44.0 ⫾ 5.2 48.3 ⫾ 7.3
of NK-sensitive YAC-1 tumor target cells in 4-h 51Cr release as-
NK cell CD3⫺DX5⫹ 3.5 ⫾ 1.0 3.3 ⫾ 0.8
DX5⫹CD62L⫹ 2.3 ⫾ 0.9 2.1 ⫾ 0.4 says. As shown in Fig. 3, NK cells from NKLAM⫺/⫺ mice had
CD3⫺NK1.1⫹ 1.6 ⫾ 0.3 1.5 ⫾ 0.4 significantly lower spontaneous NK activity in vitro compared
NKp46 90.9 ⫾ 1.5 92.4 ⫾ 2.5 with NKLAM⫹/⫺ heterozygous mice or WT mice. This represents
NKG2D 87.7 ⫾ 5.0 90.2 ⫾ 3.8 a 60% reduction in NK killing of YAC-1 targets by NKLAM-
NKG2A/C/E 48.3 ⫾ 1.3 47.0 ⫾ 1.8
Ly49A 4.5 ⫾ 1.3 4.0 ⫾ 0.9 deficient NK cells.
Ly49C/I 34.8 ⫾ 1.9 35.9 ⫾ 2.4 NKLAM expression is greatly increased in NK cells after treat-
Ly49D 49.8 ⫾ 2.1 52.4 ⫾ 3.0 ment with cytokines that enhance NK killing, such as IL-2 and
Ly49H 67.8 ⫾ 2.0 71.0 ⫾ 0.5 IFN. We therefore compared the cytotoxic activity of resting and
Ly49I 60.0 ⫾ 4.3 60.3 ⫾ 2.3
in vitro IL-2-stimulated NK cells from NKLAM⫺/⫺ and WT mice.
KLRG1 33.5 ⫾ 3.7 33.1 ⫾ 0.5
Total spleen cells (⫻106) 84.0 ⫾ 7.8 87.5 ⫾ 9.2 As seen previously, resting NK cells from NKLAM-deficient mice
a
had lower NK activity than cells from WT littermates. Interest-
Results represent the percentage of spleen cells positive for the indicated mark-
ers by flow cytometry. The mean ⫾ SD of eight individual mice per genotype is ingly, IL-2 enhanced the killing activity of both NKLAM-deficient

Downloaded from www.jimmunol.org on January 26, 2011

shown. The percentage of NK cells bearing individual NK receptors is derived from
the gated CD3⫺NK1.1⫹ population. There were no significant differences in the ab-
solute number or distribution of lymphocyte or NK subsets between WT (⫹/⫹) and
NKLAM-deficient (⫺/⫺) mice.
84 ⫾ 7.8 ⫻ 106 in WT mice. The percentages of CD3⫹, CD4⫹,
and CD8⫹ T cells, CD3⫹CD25⫹ activated T cells, CD19⫹ B cells,
and CD3⫺DX5⫹ NK cells were evaluated. We also looked at the
proportion of DX5⫹ NK cells coexpressing CD62L (L-selectin). In
humans, L-selectin expression is associated with an immature NK
cell phenotype (24). In mice, L-selectin expression on NK cells
may influence the migration of NK cells to various lymphoid or-
gans (25). As shown in Table I, there were no differences in any of
the cell populations between NKLAM-deficient and WT mice.
Analysis of the number of CD3⫺NK1.1⫹ NK cells and the NK
receptor repertoire was also performed. NKLAM-deficient and
WT mice had equivalent levels of CD3⫺NK1.1⫹ cells (Table I).
Examination of gated CD3⫺NK1.1⫹ splenic NK cells for NK in-
hibitory and activating receptors showed no differences in the ex-
pression of NKp46, NKG2D, NKG2DA/C/E, Ly49A, Ly49C/I,
Ly49D, Ly49H, Ly49I, and KLRG1 between KO and WT NK
cells (Table I and Fig. 2). Therefore, the lack of NKLAM expres-
sion does not significantly alter the numbers or distribution of NK
populations or other lymphocyte subsets in the spleen.
Maturation of NK cells was first shown to be accompanied by an
up-regulation of CD11b (Mac-1) expression (26). It was subse-
quently demonstrated that mouse NK cells could be dissected into
four subsets based on cell surface density of CD27 and CD11b
(27–29). This developmental program begins with CD11blow
CD27low NK cells and progresses through CD11blowCD27high,
CD11bhighCD27high, and CD11bhighCD27low. The distribution of
these subsets in CD3⫺NK1.1⫹ NK cells from the spleens of
NKLAM-deficient and WT mice was evaluated. As shown in Fig.
2, NKLAM KO and WT mice have similar levels of these subsets.
In addition, as described, the percentage of NK cells expressing
Ly49C/I in the CD11blow subsets of both KO and WT mice was
much lower (8%) than in the CD11bhigh subsets (30%) (data not
shown). Overall, these results suggest no NK maturational alter-
ations in NKLAM-deficient mice.
NKLAM-deficient mice display lower splenic NK activity in vitro
Previous studies have shown that NKLAM is located in NK cy-
tolytic granule membranes (9). In addition, treatment of NK cells
and WT NK cells. However, the level of killing by IL-2-stimulated
NK cells from NKLAM⫺/⫺ mice was still much less than the
killing mediated by IL-2-stimulated WT NK cells (Fig. 4). Cumu-
latively, these results indicate that NKLAM plays a role in NK
cytotoxicity in vitro. However, it is not absolute in that there is
residual killing displayed by NKLAM-deficient NK cells, which
can be enhanced by exposure to IL-2. This suggests that NKLAM
is one of multiple factors contributing to the total amount of cy-
totoxicity mediated by DX5⫹ NK cells.
NKLAM-deficient NK cells display normal granule exocytosis
and granule contents
Since NKLAM is found in the membrane of cytolytic granules, it
is possible that it plays a role in the migration and/or fusion of
granules to the cell membrane during the process of granule exo-
cytosis when NK cells are activated to kill. NKLAM levels are low
in resting NK cells and unlike other granule membrane proteins
like CD107a (LAMP-1), there is little to no preformed NKLAM in
cytolytic granules (9, 30). Upon target cell stimulation, NKLAM
levels increase. To determine whether NKLAM plays a role in
granule exocytosis, spleen cells and purified NK cells from
NKLAM⫺/⫺ and WT mice were treated with PMA and calcium
ionophore A23187 to trigger granule exocytosis. Four hours later,
the supernatants were monitored for ␤-glucuronidase, a soluble
component of the granules and a measure of granule release. Total
levels of ␤-glucuronidase in the cells were determined by deter-
gent (Triton X-100) lysis. As shown in Fig. 5, the total and re-
leased amounts of ␤-glucuronidase in NKLAM-deficient and WT
cells were comparable. This indicates that NKLAM is likely not
required for granule exocytosis and/or release of soluble granule
components.
Granule exocytosis-mediated killing of tumor targets requires
functional granules containing perforin and granzymes. If
NKLAM alters the expression and/or function of these proteins, it
could alter the killing activity of NK cells. Therefore, the levels of
perforin, granzyme B and the granule membrane protein CD107a
in NKLAM⫺/⫺ and WT NK cells were evaluated. A representative
immunoblot is shown in Fig. 6. The amount of perforin, granzyme
B, and CD107a was quantitated by densitometry and normalized to
the levels of ␤-actin. We found no significant differences in per-
forin, granzyme B, or CD107a expression between NKLAM⫺/⫺
and WT NK cells, suggesting that NKLAM does not alter the
The Journal of Immunology 6917

Downloaded from www.jimmunol.org on January 26, 2011

FIGURE 2. Normal NK cell subsets and receptor repertoire in NKLAM-deficient mice. Representative flow cytometry results from eight independent
experiments are shown. Cells isolated from the spleens of NKLAM-deficient (KO) and WT mice were stained for the markers indicated. A, Electronically
gated NK cells (CD3⫺NK1.1⫹, left panel) were evaluated for expression of the NK maturation markers CD27 and CD11b (middle panel). The dot plots
indicate the percentage of each NK subset. The right panel depicts the intracellular staining levels of granzyme B and perforin in electronically gated
CD3⫺NK1.1⫹ cells from purified NK cells stimulated with IL-2 for 3 days in vitro from KO and WT mice. B and C, The histogram plots represent the
expression levels of the indicated NK activating and inhibitory receptors electronically gated on CD3⫺NK1.1⫹ spleen cells from KO and WT mice.
Numbers indicate the percentage of positive cells for each subset.
6918
FIGURE 3. In vitro splenic NK activity is lower in NKLAM-deficient
mice. DX5⫹ NK cells from the spleens of NKLAM KO (⫺/⫺), heterozy-
gous (⫹/⫺), and WT (⫹/⫹) mice were isolated by magnetic bead column
purification. These cells were assessed for cytolytic activity against 51Cr-
labeled YAC-1 target cells in 4-h killing assays at the E:T ratios shown.
Results are expressed as the percent specific lysis of YAC-1 cells. This
represents one of six experiments with similar results. SDs of triplicate
values in each experiment did not exceed 10%.
expression of these proteins. However, we cannot rule out the pos-
sibility that there may be functional differences in the proteins
between these mice. NKLAM was undetectable in KO cells.
Intracellular levels of perforin and granzyme B were also ana-
lyzed by flow cytometry. Fig. 2A (third panel) depicts the intra-
cellular expression of these proteins in NKLAM KO- and WT-
purified NK cells stimulated for 3 days with IL-2. Cells from
NKLAM KO and WT mice have comparable levels of granzyme
B and perforin; ⬎97% of the CD3⫺NK1.1⫹ NK cells express
granzyme B and ⬎70% express perforin. Similar results were seen
in three independent experiments. Total splenocytes from
NKLAM-deficient and WT mice were also cultured for 3 days in
IL-2. Granzyme B and perforin levels in the KO and WT
CD3⫺NK1.1⫹ cells in these cultures were also comparable, with
⬎97% of the cells expressing granzyme B and ⬃45% expressing
perforin (data not shown). The flow cytometry results correlate
with the protein levels seen in immunoblot analysis. Consistent
with studies by Fehninger et al. (8), granzyme B and perforin were
undetectable in purified, unstimulated NK cells.
NKLAM-deficient mice produce less IFN-␥ upon target
stimulation in vitro
In addition to cytotoxic activity against YAC-1 tumor cells, we
compared the production of IFN-␥ by NKLAM-deficient and WT
FIGURE 4. IL-2 partially restores the killing activity of NKLAM-defi-
cient NK cells in vitro. DX5⫹ NK cells from the spleens of NKLAM KO
(⫺/⫺) and WT (⫹/⫹) mice were isolated by magnetic bead column pu-
rification. Cells were then cultured at 5 ⫻ 106 cells/ml in RPMI 1640
containing 10% FBS, 1% glutamine, and 5 ⫻ 10⫺5 M 2-ME for 3 days
with and without 500 U/ml rIL-2. These cells were assessed for cytolytic
activity against 51Cr-labeled YAC-1 target cells in 4-h killing assays at the
E:T ratios shown. Error bars represent the SDs from five independent
experiments.
ENHANCED TUMOR METASTASIS IN NKLAM-DEFICIENT MICE
FIGURE 5. NK cells and spleen cells from NKLAM KO and WT mice
display comparable levels of granule exocytosis. Granule exocytosis was
monitored by the release of ␤-glucuronidase using the fluorogenic substrate
4-methylumbelliferyl-␤-D-glucuronide. Four ⫻ 105 spleen cells or 2.5 ⫻
Downloaded from www.jimmunol.org on January 26, 2011
105 purified NK cells from NKLAM KO or WT mice were incubated with
0.025 ml of RPMI 1640 containing 4% FBS in 96-well plates. PMA (40
ng/ml) plus calcium ionophore A23187 (2 ␮g/ml) (P ⫹ I) was used to
induce granule exocytosis. Total release was obtained by incubation of
cells with 1% Triton X-100. After 4 h at 37oC, 0.005 ml of supernatant was
added to triplicate wells of 96-well black microtiter plates and mixed with
0.05 ml of ␤-glucuronidase substrate. Plates were incubated for 10 min at
37oC. The reaction was stopped by the addition of 0.175 ml of 0.25 M
sodium tetraborate. The fluorescence intensity was measured in a Wallac
Victor 1420 multilabel counter. Error bars represent the SEM from five
independent experiments. unstim, Unstimulated.
NK cells after target cell stimulation. Freshly isolated, purified,
unstimulated NK cells or NK cells stimulated with IL-2 for 3 days
in vitro were cocultured with YAC-1 target cells at E:T ratios of
FIGURE 6. Immunoblot analysis of granule proteins in NKLAM-defi-
cient and WT mice. Purified NK cells from the spleens of NKLAM-defi-
cient KO and WT mice were stimulated with IL-2 (500 U/ml) in vitro for
3 days. Cell lysates were then prepared as described and subjected to West-
ern blot analysis. The amount of perforin, granzyme B and CD107a from
four independent experiments was quantitated by densitometry and nor-
malized to the levels of ␤-actin. Levels of expression between KO and WT
cells varied by ⬍25% for each protein and were not significantly different.
NKLAM protein was undetectable in KO cells.
The Journal of Immunology
FIGURE 7. NKLAM-deficient NK cells secrete less IFN-␥ than WT
NK cells upon target cell stimulation in vitro. Freshly isolated, purified,
unstimulated NK cells (unstimulated) or NK cells stimulated with IL-2 for
3 days in vitro (IL-2) from NKLAM-deficient (KO) and WT mice were
cocultured with YAC-1 target cells at E:T ratios of 25:1 and 10:1, respec-
tively. After 4 h of incubation, levels of IFN-␥ in the supernatants were
quantitated by ELISA and expressed as pg/ml. Results are the mean ⫾ SD
of four independent experiments.
25:1 and 10:1, respectively. After 4 h of incubation, levels of
IFN-␥ in the supernatants were quantitated by ELISA. As shown
in Fig. 7, freshly isolated NK cells from WT mice produced more
IFN-␥ than NK cells from NKLAM KO mice after coculture with
YAC-1 (93 pg/ml vs 64 pg/ml). The difference in IFN-␥ produc-
tion between WT and KO NK cells was even more dramatic when
cells were first cultured in IL-2 for 3 days before the 4 h YAC-1
costimulation (1399 pg/ml vs 561 pg/ml).
NK activity in NKLAM-deficient mice is stimulated in vivo by
poly(I:C)
NK activity is enhanced by in vivo exposure to poly(I:C). Twenty-
four hours after i.p injection of mice with poly(I:C) or PBS, spleen
cells or purified NK cells from NKLAM-deficient and WT mice
were assayed for NK activity in vitro. Poly(I:C) increased the kill-
ing activity of splenocytes and purified NK cells from both
NKLAM KO and WT mice to the same extent (Fig. 8).
Enhanced lung metastases in NKLAM-deficient mice
The B16 melanoma model of experimental lung metastasis has
been widely used in studies of NK function (31–33). In these ex-
periments, NKLAM⫺/⫺ and WT mice were injected i.v. with
300,000 B16 melanoma cells. Fifteen days later, black metastatic
melanoma colonies in the lungs were counted. NKLAM-deficient
FIGURE 8. NK activity in NKLAM-deficient mice is stimulated in vivo
by poly(I:C) (p(I:C)). NKLAM-deficient (KO) and WT mice were injected
i.p. with poly(I:C) or PBS. After 24 h, spleen cells or purified NK cells
were assayed for NK activity in vitro using YAC-1 target cells in 4-h
killing assays at E:T ratios of 100:1 and 20:1, respectively. Results are
expressed as the percent specific lysis of YAC-1 cells. This represents one
of four experiments with similar results. SDs of triplicate values in each
experiment did not exceed 10%.
6919
Downloaded from www.jimmunol.org on January 26, 2011
FIGURE 9. NKLAM-deficient mice exhibit greater pulmonary metas-
tasis of B16 melanoma cells than WT mice. NKLAM-deficient KO (⫺/⫺)
and WT (⫹/⫹) mice were injected i.v. through the lateral tail vein with
300,000 B16 melanoma cells. Fifteen days later, lungs were harvested,
fixed, and the individual lobes dissected. Black metastatic melanoma col-
onies were counted on each lung lobe surface using a dissecting micro-
scope. Colonies were counted in a blinded fashion. A, Photograph shows
representative lungs from two WT (⫹/⫹) and two NKLAM-deficient KO
(⫺/⫺) mice. B, Lung tumors were quantitated in individual NKLAM KO
(⫺/⫺) and WT (⫹/⫹) male and female mice. Results indicate the mean
number of lung colonies. Error bars indicate the SEM (n ⱖ 8). The sta-
tistical significance is p ⬍ 0.001 using the Student t test.
mice had substantially higher numbers of lung metastases com-
pared with WT controls (mean 92 nodules vs 32, p ⬍ 0.01; Fig. 9).
Since a gender difference in response to B16 melanoma was re-
ported in NK receptor 2B4⫺/⫺ mice, tumor metastasis in both
NKLAM⫺/⫺ and ⫹/⫹ males and females was evaluated (34). We
observed no gender differences; both male and female
NKLAM⫺/⫺ mice showed greater numbers of tumors than their
WT gender-matched controls (Fig. 9B). In addition to having more
metastases, both male and female NKLAM-deficient mice had
larger tumor nodules than WT mice (Fig. 9A). These results indi-
cate that NKLAM-deficient mice have diminished capacity to con-
trol tumor metastases in vivo and support the role for NKLAM in
NK function in vivo.
Discussion
This report describes the generation and initial characterization of
mice with a targeted genetic deletion of NKLAM. NKLAM is a
RING-IBR-RING finger-containing transmembrane protein found
in NK cytolytic granules. It is weakly expressed in resting NK
cells; upon target cell stimulation or after incubation with cyto-
kines that enhance NK killing, NKLAM mRNA levels in NK cells
increase and protein is synthesized and becomes targeted to cyto-
plasmic granule membranes. Previous studies have shown a role
for NKLAM in perforin/granzyme-mediated cytolysis. Treatment
of NK cells with NKLAM antisense oligonucleotides inhibits their
killing activity, indicating that NKLAM participates in the cyto-
toxic process.
To further evaluate the role of NKLAM in NK function in vivo,
NKLAM-deficient mice were generated. These mice were initially
6920 ENHANCED TUMOR METASTASIS IN NKLAM-DEFICIENT MICE
129Sv/Ev ⫻ C57BL/6 and have been backcrossed for 11 genera-
tions onto a B6 background. During this process, we observed no
significant differences in the results obtained from non-back-
crossed and fully backcrossed NKLAM homozygous KO mice.
NKLAM-deficient mice were monitored for ⬎18 mo. Overall, the
mice appeared normal and were fertile. Histological examination
of the major organs of NKLAM⫺/⫺ and ⫹/⫺ mice revealed no
overt abnormalities by light microscopy. NKLAM⫺/⫺ mice devel-
oped normally and there was no obvious difference in gross mor-
phology, size, or body weight between WT and NKLAM-deficient
mice during the 18-mo period.
We also observed no differences in the numbers or distribution
of lymphoid populations in the spleens of NKLAM⫺/⫺ mice com-
pared with WT animals. The numbers of CD3⫹, CD4⫹, and CD8⫹
T cells, CD3⫹CD25⫹ activated T cells, CD19⫹ B cells,
CD3⫺DX5⫹, and CD3⫺NK1.1⫹ NK cells in both groups of mice
were identical. We also looked at the proportion of DX5⫹ NK cells
coexpressing CD62L (L-selectin). In humans, L-selectin expres-
sion is associated with an immature NK cell phenotype (24). In
mice, L-selectin expression on NK cells may influence the migra-
tion of NK cells to various lymphoid organs (25). Again, there was
no significant difference in the proportion of DX5⫹L-selectin⫹
cells in NKLAM⫹/⫹ and ⫺/⫺ mice. We also saw no maturational
differences between NKLAM KO and WT spleen NK cells based
on the levels of coexpression of CD27 and CD11b as described
previously (26 –29). Analysis of NK inhibitory and activating re-
ceptor expression by splenic NK cells in NKLAM-deficient and
WT mice also showed no differences. Evaluation of NK popula-
tions in other organs in NKLAM-deficient mice, including bone
marrow, lymph node, liver, and lungs, is currently in progress.
Although NKLAM-deficient mice have normal numbers of
splenic NK cells and a normal NK receptor repertoire, these cells
display significantly less cytotoxic activity against YAC-1 target
cells in vitro compared with heterozygous or WT control cells.
This demonstrates unequivocally that NKLAM plays a role in NK-
mediated cytotoxicity. However, it also indicates that NKLAM is
not absolutely required for all NK-mediated killing. Similarly,
when purified NK cells were stimulated with IL-2 in vitro, killing
of both NKLAM⫺/⫺ and ⫹/⫹ NK cells was increased. However,
the cytotoxic activity mediated by NKLAM-deficient NK cells
never reached the levels of WT NK cells. This indicates that IL-2
enhances the killing activity of NKLAM-deficient NK cells. It also
suggests that there is an NKLAM-dependent and NKLAM-inde-
pendent pathway used by NK cells. This is further supported by the
finding that poly(I:C) treatment in vivo enhances NK activity in
both NKLAM-deficient and WT mice to the same extent. Because
the analysis of perforin-deficient mice led to the delineation of a
perforin-independent, Fas-dependent mechanism of cytotoxicity,
studies of NKLAM-deficient mice may similarly unveil alternative
killing pathways (35).
NKLAM is found in the membrane of cytolytic granules, sug-
gesting that it might play a role in the migration and/or fusion of
granules to the cell membrane during the process of granule exo-
cytosis when NK cells are activated to kill. However, our results
indicate comparable release of the soluble granule protein ␤-glu-
curonidase by both NKLAM⫺/⫺ and WT spleen cells and purified
NK cells after PMA and calcium ionophore treatment. Experi-
ments are planned to evaluate the potential role of NKLAM in
induction of granule exocytosis by other stimuli.
We also found no differences in expression of the granule pro-
teins perforin, granzyme B, or CD107a (LAMP-1) by flow cytom-
etry and/or immunoblot analysis between NKLAM⫺/⫺ and WT
NK cells, indicating that NKLAM does not alter the expression of
these proteins. Overall, these results suggest that NKLAM is not
involved in granule protein expression and the process of granule
exocytosis. However, we cannot rule out the possibility that
NKLAM deficiency may lead to an alteration in other granule
protein expression and/or function.
In addition to decreased cytotoxic activity against YAC-1 tumor
cells mediated by NK cells from NKLAM-deficient mice, these
cells also secrete less IFN-␥ than WT NK cells after target cell
stimulation. It has been suggested that IFN-␥ is released both syn-
aptically and multidirectionally by NK cells (36). Studies are in
progress to determine whether NKLAM is involved in IFN-␥ pro-
duction and/or release by either of these two mechanisms.
We used the B16 melanoma model to evaluate the potential role
of NKLAM in tumor metastasis in vivo. B16 melanoma cells ex-
press low amounts of MHC class I and are readily killed by NK
cells. It has been well established that inhibition of NK activity by
various methods in vivo is associated with increased B16 pulmo-
nary metastasis. More recently, Kim et al. (37) demonstrated that
NK cell-deficient transgenic mice have more than a 60-fold greater
number of B16 lung metastases compared with WT mice, con-
firming a predominant role of NK cells in B16 lung metastasis
following i.v. injection. A gender difference in response to B16
Downloaded from www.jimmunol.org on January 26, 2011
melanoma was reported in mice deficient in the NK receptor 2B4
(34). We found that NKLAM-deficient mice have significantly
greater numbers of B16 lung nodules than WT mice (between 3-
and 4-fold greater).We observed no gender differences; both male
and female NKLAM⫺/⫺ mice showed greater numbers of tumors
than their WT gender-matched controls. In addition to having more
metastases, both male and female NKLAM-deficient mice had
larger tumor nodules than WT mice. These results indicate that
NKLAM-deficient mice have diminished capacity to control tumor
metastases in vivo and support the role for NKLAM in NK func-
tion in vivo. Again, these results suggest a partial rather than ab-
solute defect in NK activity in NKLAM-deficient NK cells.
These studies point to an important role for NKLAM in NK cell
function in vivo and in vitro. These NKLAM-deficient mice are a
tool for characterizing the function of NKLAM and its role in
various disease processes. It also opens up new investigations into
alternate, NKLAM-independent mechanisms used by NK cells to
control tumors and viruses in vitro and in vivo.
Disclosures
The authors have no financial conflict of interest.
References
1. Herberman, R. B., and J. R. Ortaldo. 1981. Natural killer cells: their role in
defenses against disease. Science 214: 24 –30.
2. Peters, P. J., J. Borst, V. Oorschot, M. Fukuda, O. Krahenbuhl, J. Tschopp,
J. W. Slot, and H. J. Geuze. 1991. Cytotoxic T lymphocyte granules are secretory
lysosomes, containing both perforin and granzymes. J. Exp. Med. 173:
1099 –1109.
3. Russell, J. H., and T. J. Ley. 2002. Lymphocyte-mediated cytotoxicity. Annu.
Rev. Immunol. 20: 323–370.
4. Stinchcombe, J. C., and G. M. Griffiths. 2007. Secretory mechanisms in cell-
mediated cytotoxicity. Annu. Rev. Cell Dev. Biol. 23: 495–517.
5. Trinchieri, G., D. Santoli, and H. Koprowski. 1978. Spontaneous cell-mediated
cytotoxicity in humans: role of interferon and immunoglobulins. J. Immunol. 120:
1849 –1855.
6. Kuribayashi, K., S. Gillis, D. E. Kern, and C. S. Henney. 1981. Murine NK cell
cultures: effects of interleukin-2 and interferon on cell growth and cytotoxic re-
activity. J. Immunol. 126: 2321–2327.
7. Kornbluth, J., and R. G. Hoover. 1988. Changes in gene expression associated
with IFN-␤ and IL-2 induced augmentation of human natural killer cell function.
J. Immunol. 141: 3234 –3240.
8. Fehninger, T. A., S. F. Cai, X. Cao, A. J. Bredemeyer, R. M. Presti, A. R. French,
and T. J. Ley. 2007. Acquisition of murine NK cell cytotoxicity requires trans-
lation of a pre-existing pool of granzyme B and perforin mRNAs. Immunity 26:
798 – 811.
9. Kozlowski, M., J. Schorey, T. Portis, V. Grigoriev, and J. Kornbluth. 1999. NK
lytic-associated molecule: a novel gene selectively expressed in cells with cyto-
lytic function. J. Immunol. 163: 1775–1785.
The Journal of Immunology
10. Portis, T., J. Anderson, A. Esposito, and J. Kornbluth. 2000. Gene structure of
human and mouse NKLAM, a gene associated with cellular cytotoxicity. Immu-
nogenetics 51: 546 –555.
11. van der Reijden, B. A., C. A. Erpelinck-Verschueren, B. Lowenberg, and
J. H. Jansen. 1999. TRIADs: a new class of proteins with a novel cysteine-rich
signature. Protein Sci. 8: 1557–1561.
12. Eisenhaber, B., N. Chumak, F. Eisenhaber, and M-T. Hauser. 2007. The ring
between ring fingers (RBR) protein family. Genome Biol. 8: 209.
13. Capili, A. D., E. L. Edghill, K. Wu, and K. L. B. Borden. 2004. Structure of the
C-terminal RING finger from a RING-IBR-RING/TRIAD motif reveals a novel
zinc-binding domain distinct from a RING. J. Mol. Biol. 340: 1117–1129.
14. Fang, S., K. L. Lorick, J. P. Jensen, and A. M. Weissman. 2003. RING finger
ubiquitin protein ligases: implications for tumorigenesis, metastasis and for mo-
lecular targets in cancer. Semin. Cancer Biol. 13: 5–17.
15. Ardley, H. C., and P. A. Robinson. 2005. E3 ubiquitin ligases. Essays Biochem.
41: 15–30.
16. Liu, Y. C. 2004. Ubiquitin ligases and the immune response. Annu. Rev. Immu-
nol. 22: 81–127.
17. Anandasabapathy, N., G. S. Ford, D. Bloom, C. Holness, V. Paragas, C. Seroogy,
H. Skrenta, M. Hollenhorst, C. G. Fathman, and L. Soares. 2003. GRAIL: an E3
ubiquitin ligase that inhibits cytokine gene transcription is expressed in anergic
CD4⫹ T cells. Immunity 18: 535–547.
18. Chuang T.-H., and R. J. Ulevitch. 2004. Triad3A, an E3 ubiquitin-protein ligase
regulating Toll-like receptors. Nat. Immunol. 5: 495–502.
19. Fortier, J. M., and J. Kornbluth. 2006. NK lytic-associated molecule, involved in
NK cytotoxic function, is an E3 ligase. J. Immunol. 176: 6454 – 6463.
20. Brown, M. G., A. A. Scalzo, L. R. Stone, P. Y. Clark, Y. Du, B. Palanca, and
W. M. Yokoyama. 2001. Natural killer gene complex (Nkc) allelic variability in
inbred mice: evidence for Nkc haplotypes. Immunogenetics 53: 584 –591.
21. Arase, H., T. Saito, J. H. Phillips, and L. L. Lanier. 2001. The mouse NK cell-
associated antigen recognized by DX5 monoclonal antibody is CD49b (␣2 inte-
grin, very late antigen-2). J. Immunol. 167: 1141–1144.
22. Taffs, R., M. Sitkovsky, and H. Takayama. 1998. Granule enzyme exocytosis
assay for cytotoxic T lymphocyte activation. In Current Protocols in Immunol-
ogy, Vol. 1. J. E. Coligan, A. M. Kruisbeek, D. M. Margulies, E. M. Shevach, and
W. Strober, eds. Wiley, New York, pp. 3.16.1–3.16.9.
23. Chen, J. W., T. L. Murphy, M. C. Willington, I. Pastan, and J. T. August. 1985.
Identification of two lysosomal membrane glycoproteins. J. Cell Biol. 101:
85–95.
24. Frey, M., N. B. Packianathan, T. A. Fehninger, M. E. Ross, W. C. Wang,
C. C. Stewart, M. A. Caligiuri, and S. S. Evans. 1998. Differential expression and
6921
function of L-selectin on CD56bright and CD56dim natural killer cell subsets.
J. Immunol. 161: 400 – 408.
25. Chen, S., H. Kawashima, J. B. Lowe, L. L. Lanier, and M. Fukuda. 2005. Sup-
pression of tumor formation in lymph nodes by L-selectin-mediated natural killer
cell recruitment. J. Exp. Med. 202: 1679 –1689.
26. Kim, S., K. Iizuka, H. P. Kang, A. Dokun, A. R. French, S. Greco, and
W. M. Yokoyama. 2002. In vivo developmental stages of murine natural killer
cell maturation. Nat. Immunol. 3: 523–528.
27. Hayakawa, Y., and M. J. Smyth. 2006. CD27 dissects mature NK cells into two
subsets with distinct responsiveness and migratory capacity. J. Immunol. 176:
1517–1524.
28. Hayakawa, Y., S. V. Watt, K. Takeda, and M. J. Smyth. 2008. Distinct receptor
repertoire formation in mouse NK cell subsets regulated by MHC class I expres-
sion. J. Leukocyte Biol. 83: 106 –111.
29. Chiossone, L., J. Chaix, C. Roth, E. Vivier, and T. Walzer. 2009. Maturation of
mouse NK cells is a 4-stage developmental program. Blood 113: 5488 –5496.
30. Burkhardt, J. K., S. Hester, C. K. Lapham, and Y. Argon. 1990. The lytic granules
of natural killer cells are dual-function organelles combining secretory and pre-
lysosomal compartments. J. Cell Biol. 111: 2327–2340.
31. Hanna, N., and R. C. Burton. 1981. Definitive evidence that natural killer (NK)
cells inhibit experimental tumor metastasis in vivo. J. Immunol. 127: 1754.–1758.
32. Gorelik, E., R. H. Wiltrout, K. Okumura, S. Habu, and R. B. Herberman. 1982.
Role of NK cells in the control of metastatic spread and growth of tumor cells in
mice. Int. J. Cancer 30: 107–112.
33. Warner, J. F., and G. Dennert. 1982. Effect of a cloned cell line with NK activity
on bone marrow transplants, tumour development and metastasis in vivo. Nature
300: 31–34.
34. Vaidya, S. V., S. E. Stepp, M. E. McNerney, J. Lee, M. Bennett, K. Lee,
Downloaded from www.jimmunol.org on January 26, 2011
C. L. Stewart, V. Kumar, and P. A. Mathew. 2005. Targeted disruption of the 2B4
gene in mice reveals an in vivo role of 2B4 (CD244) in the rejection of B16
melanoma cells. J. Immunol. 174: 800 – 807.
35. Kagi, D. B. Ledermann, K. Burki, P. Seiler, B. Odermatt, K. J. Olsen,
E. R. Podack, R. M. Zinkernagel, and H. Hengartner. 1994. Cytotoxicity medi-
ated by T cells and natural killer cells is greatly impaired in perforin-deficient
mice. Nature 369: 31–37.
36. Huse, M., E. J. Quann, and M. M. Davis. 2008. Shouts, whispers and the kiss of
death: directional secretion in T cells. Nat. Immunol. 9: 1105–1111.
37. Kim, S., K. Iizuka, H. L. Aguila, I. L. Weissman, and W. M. Yokoyama. 2000.
In vivo natural killer cell activities revealed by natural killer cell-deficient mice.
Proc. Natl. Acad. Sci. USA 97: 2731–2736.